CN117150356A - Service policy adjustment method and device, storage medium and electronic equipment - Google Patents
Service policy adjustment method and device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN117150356A CN117150356A CN202311067120.9A CN202311067120A CN117150356A CN 117150356 A CN117150356 A CN 117150356A CN 202311067120 A CN202311067120 A CN 202311067120A CN 117150356 A CN117150356 A CN 117150356A
- Authority
- CN
- China
- Prior art keywords
- target
- feature
- data
- user
- initial
- 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
- 238000000034 method Methods 0.000 title claims abstract description 69
- 238000013145 classification model Methods 0.000 claims abstract description 48
- 238000002474 experimental method Methods 0.000 claims abstract description 46
- 238000012545 processing Methods 0.000 claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 17
- 230000035945 sensitivity Effects 0.000 claims description 12
- 238000000926 separation method Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 description 20
- 230000008901 benefit Effects 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001364 causal effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
Abstract
The application discloses a method and a device for adjusting a business strategy, a storage medium and electronic equipment, wherein the method comprises the following steps: acquiring a plurality of experimental data obtained by performing online comparison experiments on initial business strategies to be adjusted of all target users in a target user set; performing data processing on the user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data; model training is carried out on the initial classification model based on each feature data in the feature data set to obtain a target classification model; determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set; and adjusting the initial service strategy at least based on each target feature class to obtain a target service strategy.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and apparatus for adjusting a service policy, a storage medium, and an electronic device.
Background
For each business scenario, various business strategies are continually introduced. Before the service policies are put in and applied, service experiments are usually carried out aiming at the service policies, and then the service policies are repeatedly adjusted according to historical experience according to service experiment effects so as to enable the finally adjusted service policies to achieve expected service effects.
However, the above-mentioned service policy adjustment method has a problem of low adjustment efficiency, and cannot adjust the service policy quickly and accurately.
Disclosure of Invention
In view of the above, the present application provides a method, a device, a storage medium and an electronic device for adjusting a service policy, which mainly aims to solve the problem that the service policy cannot be adjusted rapidly and accurately at present.
In order to solve the above problems, the present application provides a method for adjusting a service policy, including:
acquiring a plurality of experimental data obtained by performing online comparison experiments on initial business strategies to be adjusted of all target users in a target user set;
performing data processing on the user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
model training is carried out on the initial classification model based on each feature data in the feature data set to obtain a target classification model;
determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and adjusting the initial service strategy at least based on each target feature class to obtain a target service strategy.
Optionally, the data processing is performed on the user attribute information of each target user and each experimental data to obtain a feature data set including a plurality of feature data, which specifically includes:
performing barrel separation processing on the user attribute information and the experimental data to obtain a plurality of first characteristic data corresponding to each numerical characteristic category and a plurality of second characteristic data corresponding to each non-numerical characteristic category;
the feature data set is obtained based on each first feature data and each second feature data.
Optionally, before the barrel separation process, the method further includes:
respectively carrying out missing value processing on the user attribute information of each target user and each experimental data to obtain processed user attribute information and processed experimental data;
and respectively carrying out standardization processing on the processed user attribute information and the processed experimental data to obtain standardized user data information and standardized experimental data so as to be based on the standardized user attribute information and the standardized experimental data.
Optionally, determining a plurality of target feature classes from the feature classes by using the target classification model and the partial feature data corresponding to each feature class in the feature data set, specifically includes:
determining the influence degree of each feature class on the business target of the initial business strategy by utilizing the target classification model and partial feature data corresponding to each feature class in the feature data set;
and determining a plurality of target feature categories from the feature categories based on the influence degree of each feature category on the service targets of the initial service policy.
Optionally, before adjusting the target service policy, the method further includes:
determining target classification crowd insensitive to the initial service strategy based on experimental data corresponding to each target user in a target user set;
the method comprises the steps of adjusting the initial service policy at least based on each target feature category to obtain a target service policy, and specifically comprises the following steps:
and adjusting the initial service strategy based on the target classification crowd and each target feature class to obtain a target service strategy.
Optionally, the determining, based on experimental data corresponding to each target user in the target user set, a target classification crowd sensitive to the initial service policy specifically includes:
classifying each target user by a top-down classification mode by utilizing a preset target classification method based on attribute information of each target user in a target user set and experimental data corresponding to each target user to obtain a plurality of classification groups;
determining the sensitivity of each classified crowd to the initial business strategy;
and determining target classified crowd sensitive to the target business strategy from the classified crowd based on the sensitivity corresponding to the classified crowd.
In order to solve the above problems, the present application provides a device for adjusting a service policy, including:
the acquisition module is used for acquiring a plurality of experimental data obtained by carrying out online comparison experiments on initial business strategies to be adjusted of all target users in the target user set;
the processing module is used for carrying out data processing on the user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
the training module is used for carrying out model training on the initial classification model based on each characteristic data in the characteristic data set to obtain a target classification model;
the determining module is used for determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and the adjustment module is used for adjusting the initial service strategy at least based on each target characteristic category to obtain a target service strategy.
Optionally, the processing module is specifically configured to:
performing barrel separation processing on the user attribute information and the experimental data to obtain a plurality of first characteristic data corresponding to each numerical characteristic category and a plurality of second characteristic data corresponding to each non-numerical characteristic category;
the feature data set is obtained based on each first feature data and each second feature data.
In order to solve the above-mentioned problems, the present application provides a storage medium storing a computer program which, when executed by a processor, implements the steps of the business strategy adjustment method described in any one of the above.
In order to solve the above problems, the present application provides an electronic device, which at least includes a memory, and a processor, wherein the memory stores a computer program, and the processor implements the steps of the business strategy adjustment method according to any one of the above when executing the computer program on the memory.
According to the method, the device, the storage medium and the electronic equipment for adjusting the business strategy, the experimental data are obtained through on-line comparison experiments, then the experimental data and the user attribute information are processed to obtain the feature data set containing the feature data of each feature class, the target classification model is obtained through training, and then the target feature class can be quickly and accurately determined from each feature class based on the target classification model and the feature data set, so that the reason of negative business experiment is quickly and accurately determined, the initial business strategy can be quickly and accurately adjusted based on the target feature class, the speed of business strategy adjustment is improved, and the accuracy of business strategy adjustment is guaranteed.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:
FIG. 1 is a flowchart of a method for adjusting a business strategy according to an embodiment of the present application;
fig. 2 is a block diagram of a service policy adjustment device according to another embodiment of the present application;
fig. 3 is a block diagram of an electronic device according to another embodiment of the present application.
Detailed Description
Various aspects and features of the present application are described herein with reference to the accompanying drawings.
It should be understood that various modifications may be made to the embodiments of the application herein. Therefore, the above description should not be taken as limiting, but merely as exemplification of the embodiments. Other modifications within the scope and spirit of the application will occur to persons of ordinary skill in the art.
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and, together with a general description of the application given above, and the detailed description of the embodiments given below, serve to explain the principles of the application.
These and other characteristics of the application will become apparent from the following description of a preferred form of embodiment, given as a non-limiting example, with reference to the accompanying drawings.
It is also to be understood that, although the application has been described with reference to some specific examples, those skilled in the art can certainly realize many other equivalent forms of the application.
The above and other aspects, features and advantages of the present application will become more apparent in light of the following detailed description when taken in conjunction with the accompanying drawings.
Specific embodiments of the present application will be described hereinafter with reference to the accompanying drawings; however, it is to be understood that the disclosed embodiments are merely exemplary of the application, which can be embodied in various forms. Well-known and/or repeated functions and constructions are not described in detail to avoid obscuring the application in unnecessary or unnecessary detail. Therefore, specific structural and functional details disclosed herein are not intended to be limiting, but merely as a basis for the claims and as a representative basis for teaching one skilled in the art to variously employ the present application in virtually any appropriately detailed structure.
The specification may use the word "in one embodiment," "in another embodiment," "in yet another embodiment," or "in other embodiments," which may each refer to one or more of the same or different embodiments in accordance with the application.
The embodiment of the application provides a method for adjusting a service policy, which can be particularly applied to electronic equipment such as a terminal, a server and the like, or can also be a platform, a system and the like for adjusting the service policy. In this embodiment, taking an execution body as a server as an example, as shown in fig. 1, the method in this embodiment includes the following steps:
step S101, acquiring a plurality of experimental data obtained by performing online comparison experiments on initial business strategies to be adjusted of all target users in a target user set;
in this step, the online control experiment can be understood as a business control experiment, that is, an AB experiment for the initial business strategy. The business control experiment may be, for example: examples of business control experiments on user experience (color, font, interaction), business control experiments on algorithm optimization (search, advertisement, personalization, recommendation), business control experiments on product performance (response speed, throughput, stability, delay), etc.
In the specific implementation process, the server can respond to the adjustment request, perform data acquisition based on each target user in the request, obtain experimental data of each user for performing an AB experiment under an initial service policy, and simultaneously obtain user attribute information of each target user, wherein the user attribute information comprises: age, occupation, gender, income, hobbies, academic history, location, and the like. After obtaining the experimental data and the user attribute information, the experimental data and the user attribute information may be further stored in a predetermined data table.
Step S102, carrying out data processing on user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
in the specific implementation process, the server can preprocess each experimental data and each user attribute information, so as to obtain a plurality of feature data corresponding to each type of feature, namely, a feature data set.
Step S103, performing model training on the initial classification model based on each feature data in the feature data set to obtain a target classification model;
in the specific implementation process, model training can be performed by machine learning to obtain the target classification model. That is, the classification model in machine learning is used to train the preprocessed experimental data and the user attribute information to obtain a model file, namely, the target classification model is obtained.
Step S104, determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
in the specific implementation process, the shape analysis tool can be used for determining the target feature class with larger influence on the purpose of the business experiment from the feature classes, namely, determining the target feature class with larger influence on the business target of the initial business strategy, and laying a foundation for the subsequent adjustment of the initial business strategy based on the target feature class.
Step S105, adjusting the initial service policy based at least on each target feature class, to obtain a target service policy.
In the step, after the target feature class is determined, the service policy can be adjusted according to the target feature class, so that the adjustment of the service policy is more targeted, and the target service policy can be obtained quickly and accurately.
According to the method for adjusting the business strategy, the experimental data are obtained through on-line comparison experiments, then the experimental data and the user attribute information are processed to obtain the feature data set containing the feature data of each feature class, the target classification model is obtained through training, the target feature class can be quickly and accurately determined from each feature class based on the target classification model and the feature data set, the cause of the negative business experiment can be quickly and accurately determined, the initial business strategy can be quickly and accurately adjusted based on the target feature class, the speed of business strategy adjustment is improved, and the accuracy of business strategy adjustment is guaranteed.
In this embodiment, when data processing is performed on user attribute information of each target user and each piece of experimental data to obtain a feature data set including a plurality of feature data, a bucket-dividing processing manner may be specifically adopted. That is, the user attribute information and the experimental data are subjected to barrel division processing to obtain a plurality of first feature data corresponding to each numerical value type feature class and a plurality of second feature data corresponding to each non-numerical value type feature class; the feature data set is obtained based on each first feature data and each second feature data. In the specific implementation process of this embodiment, before the bucket separation process, the missing value process and the normalization process may be sequentially performed. Namely, the user attribute information of each target user and each experimental data are subjected to missing value processing respectively to obtain processed user attribute information and processed experimental data; and respectively carrying out standardization processing on the processed user attribute information and the processed experimental data to obtain standardized user data information and standardized experimental data so as to be based on the standardized user attribute information and the standardized experimental data.
Specifically, the first step: carrying out missing value processing on the experimental data and the user attribute information so as to fill the numerical characteristic data with 0 and fill the non-numerical characteristic data with-1; the numerical feature is understood to be feature data corresponding to the feature categories such as income and age. The non-numerical characteristic data specifically refers to characters and pictures, and can be obtained by combining: feature data related to feature categories such as financial, military, entertainment, and the like.
And a second step of: and respectively carrying out standardization and barrel separation processing on the numerical type features, and carrying out Label_Encoder coding on the non-numerical type features so as to obtain a feature data set containing a plurality of feature data. In the embodiment, through carrying out missing value processing, standardization processing, barrel separation processing and coding processing on the experimental data and the user attributes, each characteristic data can be obtained rapidly and accurately, a foundation is laid for rapidly and accurately locating the cause of the negative direction of the business experiment based on each characteristic data, and the method is beneficial to carrying out targeted adjustment on the business strategy based on the negative cause of the business experiment.
Based on the foregoing embodiment, still another embodiment of the present application provides a method for adjusting a service policy, where in this embodiment, when determining a plurality of target feature classes, a plurality of target feature classes may be determined according to influence of each feature class. That is, the influence degree of each feature class on the business target of the initial business strategy can be determined by using the target classification model and the partial feature data corresponding to each feature class in the feature data set. And determining a plurality of target feature categories from the feature categories based on the influence degree of each feature category on the business targets of the initial business strategy.
In this embodiment, when determining the influence degree of each feature class, 1% of feature data may be extracted from each feature class, and then input to the shape analysis tool together with the target classification model to obtain the shape value corresponding to each feature class, that is, obtain the influence degree of each feature class. The shape value reflects the influence degree of the feature class on the business target of the business strategy. Therefore, influence of each characteristic category on the purpose of the business experiment is obtained by analyzing by using the shape analysis tool.
When determining a plurality of target category characteristics from each category characteristic, specifically, the influence degree of each characteristic category can be respectively compared with a preset influence degree threshold value, and then the characteristic category which is larger than the preset influence degree threshold value is determined as the target characteristic category. The feature categories may be ranked in order of high to low according to the influence degree, and then a predetermined number of feature categories with a front ranking order may be determined as the target feature category.
In the specific implementation process of the embodiment, the target classification crowd sensitive to the initial service policy can be further determined, and then the initial service policy is adjusted by combining the target classification crowd, so that the adjustment of the service policy is more accurate. When determining the target classification crowd, the specific process is as follows: classifying each target user by a top-down classification mode by utilizing a preset target classification method based on attribute information of each target user in a target user set and experimental data corresponding to each target user to obtain a plurality of classification groups; determining the sensitivity of each classified crowd to the initial business strategy; and determining target classified crowd sensitive to the target business strategy from the classified crowd based on the sensitivity corresponding to the classified crowd. In this embodiment, the target classification method may specifically be an upliftttreeclassiffer method in an open source project CausalML library of Uber, and classify each target user by using the upliftttreeclassiffer method in the CausalML library, aiming at attribute information of each target user in the target user set and experimental data corresponding to each target user, thereby obtaining a classification crowd of a tree structure. Each node in the tree structure represents a classified crowd, and the sensitivity update score corresponding to each classified crowd can be obtained. Finally, the classified crowd with the largest sensitivity of the classified crowd can be determined as the target classified crowd.
In this embodiment, by calculating the sensitivity of each classified crowd, the target classified crowd sensitive to the initial business strategy can be accurately determined, so that the cause of the negative business experiment is further located, the location of the cause of the negative business experiment is more comprehensive, and further the subsequent adjustment of the initial business strategy is more accurate.
The method for adjusting the business strategy in the application is described below in connection with a specific application scenario, and the method is applied to a platform, which can be a strategy adjustment platform based on a flash framework. The platform mainly comprises four modules, namely a data layer, a model layer, a result layer and an application layer. The function of each module is as follows:
(1) Data layer: the method mainly comprises three parts of receiving and warehousing experiment analysis task request parameters, synchronizing business experiment data and user attribute data and performing characteristic engineering. In the practical use process, the platform side receives relevant parameters of an experiment to be analyzed (experimental data is combined with a table to facilitate interaction with a model layer and a result layer, and when the data is synchronous, a request parameter (experimental data) and user attribute information are needed to be combined.
(2) Model layer: the method mainly comprises two parts of feature analysis and strategy crowd analysis.
The feature analysis aims at helping the business find out some user features with great influence on business experiment targets, and the feature analysis can be specifically completed by means of machine learning and shape tools. Taking conversion rate as an example, the application firstly uses a classification model in machine learning to train experimental data (each characteristic data in a characteristic data set) after characteristic processing to obtain a model file. And then randomly extracting 1% of characteristic data from the training data and taking the characteristic data and the model file as inputs of shape analysis, thereby obtaining the influence degree of each characteristic (characteristic category) on the business experiment target.
The strategy crowd analysis aims at realizing automatic output of crowd strategies and obtaining sensitive crowd and insensitive crowd under the current business experiment. Specifically, the open source item CausalML of Uber may be used. Taking the conversion rate as an example, based on the attribute information of each target user in the target user set and the experimental data corresponding to each target user, classifying each target user by means of an UoliftTreeClassification method in a CausalmL library to obtain a plurality of classified groups. Further, the sensitivity corresponding to each classified crowd/policy crowd can be determined, so that sensitive crowd and insensitive crowd under the current business experiment can be obtained, and the target classified crowd can be obtained.
(3) In the result layer, mainly comprises two parts of data storage and result display. In order to improve the use experience of a user, the application stores the analysis result of the model layer in a warehouse at the layer, and aims to display the result at the front end, thereby facilitating the service personnel to acquire the relevant result of the experimental analysis.
(4) At the application layer, the initial business strategy can be adjusted according to the target feature class and the target classification crowd. For example, more resources are inclined to the crowd with higher conversion probability, so that the investment on insensitive crowd is reduced, and the benefit is maximized.
In the specific implementation process, the negative reasons of the experiment can be rapidly positioned according to the output result of the platform, the experiment decision can be rapidly made, and the benefit is obtained.
According to the application, the cause of the negative direction of the business experiment can be rapidly and accurately positioned by combining the basic attributes of the user based on a large amount of business experiment data and adopting the causal inference and machine learning technology, so that the subsequent rapid and accurate strategy adjustment according to the cause is facilitated, the efficiency of the business strategy adjustment is improved, and the accuracy of the business strategy adjustment is ensured.
Another embodiment of the present application provides a device for adjusting a service policy, as shown in fig. 2, including:
the acquisition module 11 is used for acquiring a plurality of experimental data obtained by performing online comparison experiments on initial service strategies to be adjusted of all target users in the target user set;
the processing module 12 is configured to perform data processing on user attribute information of each target user and each experimental data to obtain a feature data set including a plurality of feature data;
the training module 13 is used for carrying out model training on the initial classification model based on each characteristic data in the characteristic data set to obtain a target classification model;
a first determining module 14, configured to determine a plurality of target feature classes from the feature classes by using the target classification model and partial feature data corresponding to each feature class in the feature data set;
and the adjustment module 15 is configured to adjust the initial service policy based at least on each of the target feature categories, so as to obtain a target service policy.
In a specific implementation process of this embodiment, the processing module is specifically configured to:
performing barrel separation processing on the user attribute information and the experimental data to obtain a plurality of first characteristic data corresponding to each numerical characteristic category and a plurality of second characteristic data corresponding to each non-numerical characteristic category; the feature data set is obtained based on each first feature data and each second feature data.
In a specific implementation process of this embodiment, the device for adjusting a service policy further includes: the system comprises a missing value processing module and a standardized processing module. Wherein, the missing value processing module is used for: before barrel separation processing, respectively carrying out missing value processing on user attribute information of each target user and each experimental data to obtain processed user attribute information and processed experimental data; the standardized processing module is used for: and respectively carrying out standardization processing on the processed user attribute information and the processed experimental data to obtain standardized user data information and standardized experimental data so as to be based on the standardized user attribute information and the standardized experimental data.
In a specific implementation process of this embodiment, the first determining module is specifically configured to: determining the influence degree of each feature class on the business target of the initial business strategy by utilizing the target classification model and partial feature data corresponding to each feature class in the feature data set; and determining a plurality of target feature categories from the feature categories based on the influence degree of each feature category on the service targets of the initial service policy.
In a specific implementation process of this embodiment, the device for adjusting a service policy further includes: the second determining module is specifically configured to: before adjusting the target service strategy, determining a target classification crowd sensitive to the initial service strategy based on experimental data corresponding to each target user in a target user set; the adjusting module is specifically used for: and adjusting the initial service strategy based on the target classification crowd and each target feature class to obtain a target service strategy.
In a specific implementation process of this embodiment, the second determining module is specifically configured to: classifying each target user by a top-down classification mode by utilizing a preset target classification method based on attribute information of each target user in a target user set and experimental data corresponding to each target user to obtain a plurality of classification groups; determining the sensitivity of each classified crowd to the initial business strategy; and determining target classified crowd insensitive to the target business strategy from the classified crowd based on the sensitivity corresponding to the classified crowd.
According to the device for adjusting the business strategy, the experimental data are obtained through on-line comparison experiments, then the experimental data and the user attribute information are processed to obtain the feature data set containing the feature data of each feature class, the target classification model is obtained through training, the target feature class can be quickly and accurately determined from each feature class based on the target classification model and the feature data set, the reason of negative business experiment is quickly and accurately determined, the initial business strategy can be quickly and accurately adjusted based on the target feature class, the speed of business strategy adjustment is improved, and the accuracy of business strategy adjustment is guaranteed.
Another embodiment of the present application provides a storage medium storing a computer program which, when executed by a processor, performs the method steps of:
step one, acquiring a plurality of experimental data obtained by performing online control experiments on initial business strategies to be adjusted of all target users in a target user set;
step two, carrying out data processing on user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
thirdly, performing model training on the initial classification model based on each feature data in the feature data set to obtain a target classification model;
determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and fifthly, adjusting the initial service policy at least based on each target feature class to obtain a target service policy.
The specific implementation process of the above method steps can refer to the embodiment of the above method for adjusting any service policy, and this embodiment is not repeated here.
The storage medium in this embodiment obtains experimental data by performing an online comparison experiment, then processes the experimental data and user attribute information to obtain a feature data set containing feature data of each feature class, trains to obtain a target classification model, and then can quickly and accurately determine a target feature class from each feature class based on the target classification model and the feature data set, thereby realizing quick and accurate determination of the cause of the negative direction of the service experiment, further quickly and accurately adjusting an initial service policy based on the target feature class, improving the adjustment speed of the service policy, and ensuring the accuracy of service policy adjustment.
Another embodiment of the present application provides an electronic device, as shown in fig. 3, at least including a memory 1 and a processor 2, where the memory 1 stores a computer program, and the processor 2 implements the following method steps when executing the computer program on the memory 1:
step one, acquiring a plurality of experimental data obtained by performing online control experiments on initial business strategies to be adjusted of all target users in a target user set;
step two, carrying out data processing on user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
thirdly, performing model training on the initial classification model based on each feature data in the feature data set to obtain a target classification model;
determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and fifthly, adjusting the initial service policy at least based on each target feature class to obtain a target service policy.
The specific implementation process of the above method steps can refer to the embodiment of the above method for adjusting any service policy, and this embodiment is not repeated here.
The electronic equipment in the embodiment obtains experimental data by performing an online comparison experiment, processes the experimental data and the user attribute information to obtain a characteristic data set containing characteristic data of each characteristic category, trains to obtain a target classification model, and can quickly and accurately determine the target characteristic category from each characteristic category based on the target classification model and the characteristic data set, thereby realizing quick and accurate determination of the negative cause of the service experiment, further quickly and accurately adjusting the initial service policy based on the target characteristic category, improving the speed of service policy adjustment, and ensuring the accuracy of service policy adjustment.
The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application, the scope of which is defined by the claims. Various modifications and equivalent arrangements of this application will occur to those skilled in the art, and are intended to be within the spirit and scope of the application.
Claims (10)
1. The method for adjusting the service policy is characterized by comprising the following steps:
acquiring a plurality of experimental data obtained by performing online comparison experiments on initial business strategies to be adjusted of all target users in a target user set;
performing data processing on the user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
model training is carried out on the initial classification model based on each feature data in the feature data set to obtain a target classification model;
determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and adjusting the initial service strategy at least based on each target feature class to obtain a target service strategy.
2. The method of claim 1, wherein the data processing is performed on the user attribute information of each target user and each of the experimental data to obtain a feature data set including a plurality of feature data, and specifically includes:
performing barrel separation processing on the user attribute information and the experimental data to obtain a plurality of first characteristic data corresponding to each numerical characteristic category and a plurality of second characteristic data corresponding to each non-numerical characteristic category;
the feature data set is obtained based on each first feature data and each second feature data.
3. The method of claim 2, wherein prior to the barreling, the method further comprises:
respectively carrying out missing value processing on the user attribute information of each target user and each experimental data to obtain processed user attribute information and processed experimental data;
and respectively carrying out standardization processing on the processed user attribute information and the processed experimental data to obtain standardized user data information and standardized experimental data so as to be based on the standardized user attribute information and the standardized experimental data.
4. The method of claim 1, wherein determining a plurality of target feature classes from each of the feature classes using the target classification model and the partial feature data corresponding to each feature class in the feature data set, specifically comprises:
determining the influence degree of each feature class on the business target of the initial business strategy by utilizing the target classification model and partial feature data corresponding to each feature class in the feature data set;
and determining a plurality of target feature categories from the feature categories based on the influence degree of each feature category on the service targets of the initial service policy.
5. The method of claim 1, wherein prior to adjusting the target traffic policy, the method further comprises:
determining target classification crowd sensitive to the initial service strategy based on experimental data corresponding to each target user in a target user set;
the method comprises the steps of adjusting the initial service policy at least based on each target feature category to obtain a target service policy, and specifically comprises the following steps:
and adjusting the initial service strategy based on the target classification crowd and each target feature class to obtain a target service strategy.
6. The method of claim 5, wherein the determining the target taxonomy group sensitive to the initial business policy based on experimental data corresponding to each target user in the set of target users specifically comprises:
classifying each target user by a top-down classification mode by utilizing a preset target classification method based on attribute information of each target user in a target user set and experimental data corresponding to each target user to obtain a plurality of classification groups;
determining the sensitivity of each classified crowd to the initial business strategy;
and determining target classified crowd sensitive to the target business strategy from the classified crowd based on the sensitivity corresponding to the classified crowd.
7. A device for adjusting a service policy, comprising:
the acquisition module is used for acquiring a plurality of experimental data obtained by carrying out online comparison experiments on initial business strategies to be adjusted of all target users in the target user set;
the processing module is used for carrying out data processing on the user attribute information of each target user and each experimental data to obtain a characteristic data set containing a plurality of characteristic data;
the training module is used for carrying out model training on the initial classification model based on each characteristic data in the characteristic data set to obtain a target classification model;
the first determining module is used for determining a plurality of target feature categories from the feature categories by utilizing the target classification model and partial feature data corresponding to the feature categories in the feature data set;
and the adjustment module is used for adjusting the initial service strategy at least based on each target characteristic category to obtain a target service strategy.
8. The apparatus of claim 7, wherein the processing module is specifically configured to:
performing barrel separation processing on the user attribute information and the experimental data to obtain a plurality of first characteristic data corresponding to each numerical characteristic category and a plurality of second characteristic data corresponding to each non-numerical characteristic category;
the feature data set is obtained based on each first feature data and each second feature data.
9. A storage medium storing a computer program which, when executed by a processor, implements the steps of the business strategy adjustment method of any of the preceding claims 1-6.
10. An electronic device comprising at least a memory, a processor, said memory having stored thereon a computer program, said processor, when executing the computer program on said memory, implementing the steps of the method for adjusting a traffic policy according to any of the preceding claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311067120.9A CN117150356A (en) | 2023-08-23 | 2023-08-23 | Service policy adjustment method and device, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311067120.9A CN117150356A (en) | 2023-08-23 | 2023-08-23 | Service policy adjustment method and device, storage medium and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117150356A true CN117150356A (en) | 2023-12-01 |
Family
ID=88901871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311067120.9A Pending CN117150356A (en) | 2023-08-23 | 2023-08-23 | Service policy adjustment method and device, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117150356A (en) |
-
2023
- 2023-08-23 CN CN202311067120.9A patent/CN117150356A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9449271B2 (en) | Classifying resources using a deep network | |
| US20210382937A1 (en) | Image processing method and apparatus, and storage medium | |
| CN110751224B (en) | Training method of video classification model, video classification method, device and equipment | |
| US8498950B2 (en) | System for training classifiers in multiple categories through active learning | |
| CN112819023A (en) | Sample set acquisition method and device, computer equipment and storage medium | |
| CN113011889B (en) | Account anomaly identification method, system, device, equipment and medium | |
| CN113298197B (en) | Data clustering method, device, equipment and readable storage medium | |
| CN110750523A (en) | Data annotation method, system, computer equipment and storage medium | |
| CN112966088B (en) | Unknown intention recognition method, device, equipment and storage medium | |
| CN113641797A (en) | Data processing method, device, equipment, storage medium and computer program product | |
| CN113569118A (en) | Self-media pushing method and device, computer equipment and storage medium | |
| CN116563604A (en) | End-to-end target detection model training, image target detection method and related equipment | |
| CN114444514B (en) | Semantic matching model training method, semantic matching method and related device | |
| CN117150356A (en) | Service policy adjustment method and device, storage medium and electronic equipment | |
| CN115564578B (en) | Fraud recognition model generation method | |
| CN115130536A (en) | Training method of feature extraction model, data processing method, device and equipment | |
| CN113886547A (en) | Client real-time conversation switching method and device based on artificial intelligence and electronic equipment | |
| CN113704623A (en) | Data recommendation method, device, equipment and storage medium | |
| CN109308565B (en) | Crowd performance grade identification method and device, storage medium and computer equipment | |
| CN113628077A (en) | Method for generating non-repeated examination questions, terminal and readable storage medium | |
| CN111091198A (en) | Data processing method and device | |
| CN116860952B (en) | RPA intelligent response processing method and system based on artificial intelligence | |
| CN113743443B (en) | Image evidence classification and recognition method and device | |
| EP3923154A1 (en) | Identifying and generating links between data | |
| CN110119506B (en) | Text extraction method, system, equipment and storage medium |
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 |