CN112685674A - Feature evaluation method and device influencing user retention - Google Patents
Feature evaluation method and device influencing user retention Download PDFInfo
- Publication number
- CN112685674A CN112685674A CN202011613766.9A CN202011613766A CN112685674A CN 112685674 A CN112685674 A CN 112685674A CN 202011613766 A CN202011613766 A CN 202011613766A CN 112685674 A CN112685674 A CN 112685674A
- Authority
- CN
- China
- Prior art keywords
- intervention
- variable
- account
- retention
- value
- 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
- 230000014759 maintenance of location Effects 0.000 title claims abstract description 123
- 238000011156 evaluation Methods 0.000 title description 7
- 230000001364 causal effect Effects 0.000 claims abstract description 131
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 21
- 239000000523 sample Substances 0.000 claims description 35
- 238000012549 training Methods 0.000 claims description 17
- 239000013068 control sample Substances 0.000 claims description 12
- 238000002474 experimental method Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000000717 retained effect Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses a method and a device for evaluating characteristics influencing user retention, wherein the method comprises the following steps: acquiring corresponding specified characteristic values of each account in the current platform according to the specified characteristic variables; acquiring retention information of each account, wherein the retention information is used for reflecting whether the current account is a retention account; determining an intervention variable according to the acquired specified characteristic value; and determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information. In the embodiment, the influence of each characteristic variable on the user retention is judged by a causal inference method, so that key factors influencing the user retention can be quickly analyzed, and after the key factors are obtained, the platform side can pertinently improve the factors through some customized strategies, so that the retention rate of the user is improved.
Description
Technical Field
The embodiment of the application relates to a data processing technology, in particular to a method and a device for evaluating characteristics influencing user retention.
Background
In the internet industry, a user starts to use an application within a certain period of time, and after a certain period of time, the application still continues to be used and is regarded as retention, and the retention rate is the proportion of the user to the newly added user at that time. The concept of retention can be used for analyzing the service effect of the application or the website and judging whether the application or the website can retain the user. The retention rate reflects the fact that it is a conversion rate from an unstable user in the early stage to an active user, a stable user, and a loyal user. In general, if the platform side does not take action to increase the retention rate, the retention rate of the new user will be at a lower level. Therefore, the platform side needs to take targeted measures to increase the retention rate of the user.
Disclosure of Invention
The application provides a feature evaluation method and device influencing user retention, so that key factors influencing user retention are evaluated, and therefore targeted measures are taken to increase the retention rate of a user.
In a first aspect, an embodiment of the present application provides a method for evaluating a feature affecting user retention, where the method includes:
acquiring corresponding specified characteristic values of each account in the current platform according to the specified characteristic variables;
acquiring retention information of each account, wherein the retention information is used for reflecting whether the current account is a retention account;
determining an intervention variable according to the acquired specified characteristic value;
and determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information.
In a second aspect, an embodiment of the present application further provides a feature evaluation apparatus for influencing user retention, where the apparatus includes:
the system comprises a designated characteristic value acquisition module, a characteristic value analysis module and a characteristic value analysis module, wherein the designated characteristic value acquisition module is used for acquiring corresponding designated characteristic values of accounts in a current platform according to a plurality of designated characteristic variables;
the retention information acquisition module is used for acquiring retention information of each account, and the retention information is used for reflecting whether the current account is a retention account;
the intervention variable determining module is used for determining an intervention variable according to the acquired specified characteristic value;
and the causal effect determination module is used for determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information.
In a third aspect, an embodiment of the present application further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the above method when executing the program.
In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the method described above.
The application has the following beneficial effects:
in this embodiment, after the corresponding specified feature values of the accounts in the current platform and the retention information of the accounts are acquired from the historical data, a cause-and-effect inference algorithm can be adopted to determine the cause-and-effect of a single feature on the user retention, and the key factors influencing the user retention are found according to the cause-and-effect of the features on the user retention, so that the platform side is assisted to perform targeted operation through the key factors, and the retention rate of the user can be improved.
Drawings
FIG. 1 is a flowchart of an embodiment of a method for evaluating features affecting user retention according to an embodiment of the present application;
FIG. 2 is a flowchart of an embodiment of a method for determining a causal effect of an intervention variable on user retention according to an embodiment of the present application;
FIG. 3 is a block diagram of a structure of an embodiment of a feature evaluation apparatus for influencing user retention according to a second embodiment of the present application;
fig. 4 is a schematic structural diagram of an electronic device according to a third embodiment of the present application.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the application and are not limiting of the application. It should be further noted that, for the convenience of description, only some of the structures related to the present application are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of an embodiment of a method for evaluating a feature affecting user retention according to an embodiment of the present application, where the embodiment may be applied to a server, and specifically may include the following steps:
and step 110, acquiring corresponding specified characteristic values of each account in the current platform according to the specified characteristic variables.
In an embodiment, the specified characteristic variable may be a characteristic variable specified by the platform according to prior knowledge and having a certain influence on the user, or a characteristic variable set according to an actual service requirement, which is not limited in this embodiment.
For example, for a live platform, the specified feature variables may include, but are not limited to: language features, country features, registration features, age features, gender features, device information, operator information, network information, and the like. The language features are used for describing the use language of the user; the country feature is used for describing country (region) information on user registration data; the registration feature is used for describing whether the user is registered or not; the age characteristic is used for describing the age of the user; the gender characteristic is used for describing gender information of the user; the device information is used for describing the model of the terminal (such as the model of a mobile phone) used by the user; the operator information is used for describing operator information of a network used by the user; the network information is used to describe the type of network that the user often uses.
According to the designated characteristic variable, the designated characteristic value corresponding to each account in the current platform can be collected from historical data. For example, the feature values corresponding to the language features may include chinese, english, french, german, arabic, and the like; the characteristic value corresponding to the national feature may include china, usa, uk, germany, russia, etc.; the feature values corresponding to the registered features may include registration, visitor, etc.; the feature value corresponding to the age feature may include adult, minor, etc.; the characteristic values corresponding to the gender characteristics may include male, female, etc.; the characteristic value corresponding to the device information may include a specific device model; the characteristic value corresponding to the operator information may include specific operators such as mobile, universal, telecommunication, etc.; the characteristic values corresponding to the network information may include 2g-4g, WI-FI, etc.
In one embodiment, after a series of specified feature values of each account in the current platform are collected, the specified feature values may be combined into a feature matrix, for example, the feature matrix may be represented by X, and the dimension of X may be dimX ═ n, p, where n is the number of accounts in the current platform and p is the number of specified feature values of a single account, and p may be used1，…，pnTo indicate.
For convenience of subsequent fitting, in an example, the value of the specified characteristic variable (i.e., the specified characteristic value) may be expressed by a numerical value of 0 or 1, for example, when data collection is performed, the specified characteristic value may be determined in the form of an option or a list, for example, if the specified characteristic variable is a linguistic feature, a plurality of linguistic candidates may be included under the linguistic feature, each candidate includes two options of "yes" and "no", and when the specified variable is "yes", the corresponding specified characteristic value is "1"; when selected as "no", the corresponding specified feature value is "0".
And step 120, obtaining the retention information of each account, wherein the retention information is used for reflecting whether the current account is a retention account.
In practice, the platform side may define the user retention judgment rule according to actual business requirements or experience, so as to determine the retention information of each account according to the retention judgment rule, that is, judge whether each account is a retention account. For example, if the retention judgment rule is set to be a time period, if no log-in record exists in the time period after the account is registered, the account is determined to be the attrition account, otherwise, if the account has the log-in record in the time period, the account is determined to be the retention account. For example, if an account is not logged in within 15 days after registration, it is determined that the account is not retained.
In one example, the retained information can be represented by y, which can take on the value 0 or 1 if y is presenti1, representing the final retention of user i; on the contrary, if yiAnd 0, indicates that the user i finally runs away. The y dimension is dimy ═ n, 1.
And step 130, determining an intervention variable according to the acquired specified characteristic value.
The goal of this embodiment is to analyze the impact of any single variable on retention improvement from historical data, which is the intervention variable in this embodiment, also called the primary variable, i.e., the variable of interest (i.e., to be explored) that has an impact on the target utility (i.e., user retention).
The intervention variable may be denoted by T, and may take a value of 0 or 1, representing "intervention accepted" when T is 1 and "intervention not accepted" when T is 0.
In one embodiment, step 130 may further include the steps of:
determining a target specified characteristic variable from the plurality of specified characteristic variables; and determining a target specified characteristic value from one or more specified characteristic values corresponding to the target specified characteristic variable to generate an intervention variable.
In this step, the intervention variable and the specified characteristic variable are not in an equivalent relationship, and the intervention variable may be generated according to a characteristic value in one of the specified characteristic variables. In this embodiment, the target specified characteristic variable is a specified characteristic variable of interest, and the target specified characteristic value is one of all specified characteristic values of the target specified characteristic variable.
For example, assume that the target specifies a feature variable piTo "language feature," the corresponding specified feature values include, but are not limited to: according to the different specified characteristic values of the target specified characteristic variables, the following different intervention variables (similar to one-hot encoding in form) can be generated:
T1whether or not to speak Chinese. 1 represents a spoken Chinese language and 0 represents a non-spoken Chinese language;
T2whether or not to speak English. 1 stands for English speaking and 0 stands for not.
And step 140, determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information.
In this embodiment, the user persistence is used as the target utility variable U, and the causal effect of the currently concerned intervention variable T on the target utility variable U, that is, the influence degree of the change of the intervention variable T on the target utility variable U, is determined according to a causal inference algorithm.
In one embodiment, as shown in fig. 2, step 140 may further include the steps of:
and 140-1, acquiring a first observation variable value corresponding to the account which receives the intervention and a second observation variable value corresponding to the account which does not receive the intervention according to the intervention variable.
In this step, after the intervention variable T is determined, it may be determined whether each account of the current platform belongs to an account that actually receives an intervention or an account that does not receive an intervention according to the intervention variable. The account which receives the intervention is the account which meets T-1, and the account which does not receive the intervention is the account which meets T-0. For example, if T is "whether or not to speak chinese (1 represents chinese and 0 represents no chinese)", the account that accepts the intervention is the account that speaks chinese (T ═ 1), and the account that does not accept the intervention is the account that does not speak chinese (T ═ 0).
The observed variable value (including the first observed variable value and the second observed variable value) of each account may be understood as the variable value of each account corresponding to the target utility variable in the real case, and in this embodiment, is the variable value of the target utility variable "retained by the user", i.e., "whether or not to retain". For example, for Ti1 (i.e. the account receiving the intervention), assuming that user i is eventually lost, its first observed variable value is 0; as another example, for Tj0 (i.e. an account that does not accept intervention), assuming that user j ultimately retains, its second observed variable value is 1.
And 140-2, determining a first predictive variable value of the account which receives the intervention under the condition of not receiving the intervention, and determining a second predictive variable value of the account which does not receive the intervention under the condition of receiving the intervention.
In one embodiment, the causal inference algorithm of this embodiment is based on an algorithm under the Rubin framework, and the idea of causal inference is closely related to "counterfactual inference", and for an account (T ═ 1) that receives an intervention, the causal effect of the intervention on a target utility variable is related to the value of the target utility variable under an unacceptable intervention (T ═ 0) by the account. Vice versa, for an account not accepting an intervention (T ═ 0), the causal effect of the intervention on the target utility variable is related to the value of the target utility variable under the condition that the account accepts the intervention (T ═ 1), so that the disturbance of the variable can be eliminated. However, in practice, for a single account, only one of its accepted or not accepted interventions that actually occur is typically observed, and neither is observed at the same time. Therefore, in this step, another situation that cannot be observed can be predicted to obtain a corresponding predicted variable value.
Specifically, for an account that receives an intervention, the actual condition that can be observed is the value of the target utility variable when the account receives the intervention, that is, the counterfactual condition that the account does not receive the intervention is not observed, so that the value of the target utility variable (that is, the first predictive variable value) of the account that does not receive the intervention can be predicted. Accordingly, for an account that does not receive intervention, the actual condition that can be observed is the value of the target utility variable when the account does not receive intervention, that is, the condition that the account receives intervention is not observed, and therefore the value of the target utility variable (that is, the second predictive variable value) of the account that receives intervention can be predicted.
In one embodiment, a machine model may be used to simulate counterfactual reasoning to obtain the first predicted variable value and the second predicted variable value, and step 140-2 may further include the steps of:
step 140-2-1, organize all accounts that received intervention into a set of experimental samples, and organize all accounts that did not receive intervention into a set of control samples.
In this embodiment, the account (T) that will be subject to intervention may bei1) as experimental sample, account (T) that will not accept interventioni0) as a control sample (which may also be referred to as a control sample), then all of the experimental samples may constitute a set of experimental samples and all of the control samples may constitute a set of control samples.
In one example, the experiment sample set may include, in addition to the accounts that receive the intervention, a specific feature value corresponding to each account that receives the intervention and retention information. Similarly, the experiment sample set may include, in addition to the account not to be subjected to the intervention, a designated feature value and retention information corresponding to each account not to be subjected to the intervention.
Step 140-2-2, training a first machine model based on the set of experimental samples, and training a second machine model based on the set of control samples.
In this step, the set of experimental samples is used to train a first machine model and the set of control samples is used to train a second machine model.
In one embodiment, the first machine model may be obtained as follows:
selecting a target covariate from covariates according to the intervention variable, wherein the covariate comprises: the other specified characteristic variables except the specified characteristic variable corresponding to the intervention variable in the specified characteristic variables; and training a first machine model by adopting a preset machine learning algorithm based on the variable values corresponding to the target covariates of the accounts in the experiment sample set and the retention information of the accounts.
Specifically, the experimental sample set may include intervention variables, covariates, and corresponding variable values. Covariates were introduced for the purpose of the effect-dependent interpretation of the intervention variables, with the hope that after introduction, the non-confounding assumption holds. Covariates refer to variables that are relevant to the study, in addition to intervention variables and target utility variables. In this embodiment, the covariates may include: and other specified characteristic variables except the specified characteristic variables corresponding to the intervention variables in the specified characteristic variables. It will also be appreciated that the feature matrix X may be divided into intervention variables and covariates, i.e.
Wherein the content of the first and second substances,
are covariates.
In practice, of the target specified characteristic variables corresponding to the current intervention variables, variables generated by specified characteristic values other than the target specified characteristic value also belong to covariates. For example, if T1Whether or not to say Chinese, 1 represents to say Chinese, 0 represents not to say Chinese; t is2Whether or not to speak English, 1 represents English, and 0 represents not to speak English. If the effect of T1 on the target utility variable y needs to be determined, then T2Also belonging to covariates.
After the covariates are determined, one covariate can be selected from the covariates as a target covariate, for an experiment sample set, the variable value corresponding to the target covariate of each account in the set and the retention information of the account can be used as sample data, a preset machine learning algorithm is adopted to train a first machine model corresponding to the target covariate, and the training target is to estimate the target variable value of the intervention variable under the condition of accepting the intervention by learning the association relationship between the variable value corresponding to the target covariate of each account and the retention information of the account under the condition of giving the target covariate.
In other embodiments, the target covariate may be more than one, but a plurality of, or all of the covariates, and the training target may be specifically determined according to the actual business requirement, which is not limited in this embodiment.
In one example, the predetermined machine learning algorithm may include supervised learning methods such as linear regression, random forest, support vector machine, XGBOOST, etc.
For the training method of the second machine model, similar to the training method of the first machine model, for the comparison sample set, the variable value corresponding to the target covariate of each account in the set and the retention information of the account can be used as sample data, the second machine model is trained by adopting a preset machine learning algorithm, and the trained target is to estimate the target variable value of the intervention variable under the condition of not accepting the intervention by learning the association relationship between the variable value corresponding to the target covariate of each account and the retention information of the account under the condition of giving the target covariate.
In this embodiment, the first machine model is used to estimate the value of the target variable of the account under intervention given the target covariate, and in one example, the first machine model may be defined as: mu.s1(x)＝E[U(1)|X＝x]Wherein X ═ X represents the target covariate, and U (1) represents the potential outcome of the target utility variable at T ═ 1; e [ U (1) | X ═ X]What is shown is the predicted value of the target utility variable when X ═ X and T ═ 1, i.e., given the target covariate X, when the current account is subject to intervention.
The second machine model is used to estimate target variable values for the account without accepting intervention, given the target covariates, which in one example may be defined as: mu.s0(x)＝E[U(0)|X＝x]Wherein X-X represents the target covariate and U (0) represents the potential outcome of the target utility variable when T-0; e [ U (0) | X ═ X]What is shown is the predicted value of the target utility variable when X is X and T is 0, i.e., when the current account does not accept intervention, given the target covariate X.
And 140-2-3, predicting the first prediction variable value of each account in the experiment sample set under the condition of not accepting intervention by adopting the second machine model.
In this step, for the set of experimental samples, the target variable value of each sample in the case of accepting the intervention is observable, and the target variable value of each sample in the case of not accepting the intervention is absent, so that the embodiment may use the second machine model to predict the first predicted variable value of each account in the set of experimental samples that is not accepting the intervention given the target covariate. Specifically, the variable values of the target covariates of each account in the experiment sample set may be input to the second machine model, the second machine model performs prediction processing, and the first predicted variable values corresponding to each account may be output.
And 140-2-4, predicting a second predicted variable value of each account in the control sample set under the condition of receiving the intervention by using the first machine model.
In this step, for the set of control samples, the target variable value of each sample in the case of not accepting the intervention is observable, and the target variable value of each sample in the case of accepting the intervention is absent, so that the present embodiment may use the first machine model to predict the second predicted variable value of each account in the set of control samples, which accepts the intervention in the case of the given target covariate. Specifically, the variable value of the target covariate of each account in the comparison sample set may be input to the first machine model, the first machine model may perform the prediction processing, and the second predicted variable value corresponding to each account may be output.
In one embodiment, after the second machine model predicts the first predicted variable value of each account in the experiment sample set without intervention and the second predicted variable value of each account in the control sample set with intervention, the first predicted variable value and the second predicted variable value may be subjected to error processing, for example, corresponding error terms are added to the first predicted variable value and the second predicted variable value, respectively, to facilitate data fitting and reduce errors of the model output data. The first predicted variable value of the account that accepts intervention without accepting intervention may be expressed as: u (0) ═ mu0(X) + e (0), wherein e (0) is an error term; the second predictive variable value of the account not accepting the intervention in the case of accepting the intervention may be expressed as: u (1) ═ mu1(X) +. epsilon.1, where epsilon.1 is an error term.
Step 140-3, determining a first causal effect of the account subject to intervention based on the first observed variable value and the first predicted variable value.
In this step, after obtaining the target variable values (i.e., the first observed variable value and the first predicted variable value) for each experimental sample in the set of experimental samples for both cases of intervention and non-intervention, the difference between the target variable values for the two cases can be calculated, and the difference can be used as the causal effect of the intervention variable of the sample on the target utility variable given the target covariate. In this embodiment, the difference between the first observed variable value and the first predicted variable value of the current account is calculated as the first causal effect.
For example, the first causal effect for each experimental sample can be expressed as:
wherein the content of the first and second substances,
representing a first observed variable value of a user i under the condition of accepting the intervention;
the first predicted variable value of the user i without receiving the intervention when the target intervention variable X is given and predicted by the second machine model;
representing a first causal effect of the current intervention variable on the target utility variable for user i.
In other embodiments, the first causal effect may also be expressed as an expectation of calculating a difference between the first observed variable value and the first predicted variable value for the current account, i.e., E [ D1|X＝x]。
And 140-4, determining a second causal effect of the account not accepting the intervention according to the second observed variable value and the second predicted variable value.
In this step, after obtaining the target variable values (i.e., the second predicted variable value and the second observed variable value) for each control sample in the set of control samples for both cases of intervention and non-intervention, the difference between the target variable values for the two cases can be calculated, and this difference can be used as the causal effect of the intervention variable for the sample on the target utility variable given the target covariate. In this embodiment, the difference between the second predicted variable value and the second observed variable value of the current account is calculated as the second causal effect.
For example, the second causal effect for each control sample can be expressed as:
wherein the content of the first and second substances,
representing a second observed variable value of user i without accepting intervention;
a second predicted variable value of the user i under the condition of accepting the intervention when the target intervention variable X is given and predicted by adopting the first machine model;
representing a second causal effect of the current intervention variable on the target utility variable for user i.
In other embodiments, the second causal effect may also be expressed as an expectation of calculating a difference between the second observed variable value and the second predicted variable value for the current account, i.e., E [ D0|X＝x]。
And step 140-5, determining the causal effect of the intervention variable on the user retention according to the first causal effect or the second causal effect of each account.
After the first causal effect or the second causal effect of each account is determined, the causal effect of the intervention variable on the user retention can be determined in combination with the first causal effect or the second causal effect of each account.
In one embodiment, step 140-5 may further include the steps of:
and 140-5-1, calculating the average value of all the first causal effects as the first average causal effect of the experimental sample set.
In this step, after obtaining the first causal effect of each sample (i.e. the account receiving the intervention) in the set of experimental samples, an average value of the first causal effect of each sample in the set of experimental samples may be calculated as the first average causal effect of the set of experimental samples given the target covariate, which may be understood as the average causal effect of T ═ 1 on U given the target covariate.
And 140-5-2, calculating the average value of all the second causal effects as the second average causal effect of the control sample set.
In this step, after obtaining the second causal effect of each sample in the set of control samples (i.e. the account not receiving intervention), an average value of the second causal effect of each sample in the set of control samples may be calculated as the second average causal effect of the set of control samples given the target covariate, which may be understood as the average causal effect of T ═ 0 on U given the target covariate.
It should be noted that, in step 140-5-1 and step 140-5-2, the average causal effect is determined by calculating an average value of the causal effects in the set, and in other embodiments, the average causal effect may also be determined by calculating an average expectation of the causal effects in the set, which is not limited in this embodiment.
At step 140-5-3, a trend value is determined reflecting the probability of the user accepting intervention given the target covariate.
In this embodiment, the tendency value may also be referred to as a probability that the account receives an intervention, which may be defined as e (X) ═ P (T ═ 1| X ═ X), i.e., a probability that the user receives an intervention is predicted when the target covariate X is given in a real scene, and T is a bernoulli distribution of the service parameter e (X) given X ═ i, i.e., T to Bern (e (X)).
In one embodiment, the trend value may be determined as follows:
and inputting the variable value of the target covariate into a trained probability model, and outputting the probability of the user to be intervened predicted under the target covariate as a tendency value by the probability model.
In this embodiment, a probabilistic model may be trained in advance, the training objective of which is the probability of accepting intervention given a target covariate. Then, the variable value of the current target covariate can be input into the probability model, and the probability value output by the probability model is obtained as the tendency value.
And 140-5-4, taking the tendency value as a weight, and carrying out weighted calculation on the first average causal effect and the second average causal effect to obtain the causal effect of the intervention variable on the user retention under the given target covariate.
In this step, after the trend value is obtained, the trend value may be used as a weight to perform weighted calculation on the first average causal effect and the second average causal effect, so as to obtain a causal effect of the current intervention variable on the user retention under the given target covariate, and the causal effect may be represented by the following formula:
τ(x)＝e(x)τ0(x)+(1-e(x))τ1(x)
wherein e (x) is a tendency value, τ0(x) For a second average causal effect, τ1(x) For the first average causal effect, x is the target covariate, and τ (x) is the causal effect of the prediction variables on the user's retention under a given target covariate.
In one embodiment, if multiple target covariates are specified, i.e., if there are multiple target covariates, step 140-5 may further comprise the steps of:
and calculating the average expectation of the causal effect corresponding to each target covariate, and obtaining the causal effect of the intervention variable on the user retention.
In this embodiment, if there are a plurality of target covariates, if it is determined that the causal effect of the current intervention variable on the user retention is to be performed, average expected calculation may be performed on the causal effects corresponding to the plurality of target covariates, and finally, the causal effect of the intervention variable on the user retention is obtained. In one embodiment, when
When there are a plurality of intervention variables, the embodiment may further include the following steps:
and selecting the intervention variable with the largest causal effect as a key factor influencing the retention of the user according to the causal effect of each intervention variable on the retention of the user.
In this embodiment, when a plurality of intervention variables are specified, after a causal effect of each intervention variable on the user retention is obtained, the intervention variables may be sorted according to the magnitude of the causal effect, and the sorted result may be input to a subsequent matching algorithm to help calculate the total utility, or provide a basis for manual matching.
Furthermore, one or more intervention variables with the largest causal effect can be screened out according to the sequencing result and used as key factors which have the largest influence on user retention, so that a targeted operation strategy can be formulated according to the screened intervention variables to improve the retention rate of the user.
In this embodiment, after the corresponding specified feature values of the accounts in the current platform and the retention information of the accounts are acquired from the historical data, a cause-and-effect inference algorithm can be adopted to determine the cause-and-effect of a single feature on the user retention, and the key factors influencing the user retention are found according to the cause-and-effect of the features on the user retention, so that the platform side is assisted to perform targeted operation through the key factors, and the retention rate of the user can be improved.
Example two
Fig. 3 is a block diagram of a structure of an embodiment of a feature evaluation apparatus for influencing user retention according to a second embodiment of the present application, which may include the following modules:
the designated characteristic value acquisition module 310 is configured to acquire a corresponding designated characteristic value of each account in the current platform according to the plurality of designated characteristic variables;
the retention information acquiring module 320 is configured to acquire retention information of each account, where the retention information is used to reflect whether a current account is a retention account;
an intervention variable determination module 330, configured to determine an intervention variable according to the acquired specified feature value;
a causal effect determination module 340, configured to determine a causal effect of the intervention variable on the user retention by using a causal inference algorithm in combination with the retention information.
In one embodiment, the intervention variable determination module 330 is specifically configured to:
determining a target specified characteristic variable from the plurality of specified characteristic variables;
and determining a target specified characteristic value from one or more specified characteristic values corresponding to the target specified characteristic variable to generate an intervention variable.
In one embodiment, the causal effect determination module 340 may further include the following sub-modules:
the observation variable value acquisition submodule is used for acquiring a first observation variable value corresponding to the account which receives the intervention and acquiring a second observation variable value corresponding to the account which does not receive the intervention according to the intervention variable;
the predictor variable value determination submodule is used for determining a first predictor variable value of the account which is subjected to the intervention under the condition of not receiving the intervention and determining a second predictor variable value of the account which is not subjected to the intervention under the condition of receiving the intervention;
a first causal effect determination submodule for determining a first causal effect of the account subject to intervention based on the first observed variable value and the first predicted variable value;
a second causal effect determination submodule for determining a second causal effect for the non-intervention accepting account based on the second observed variable value and the second predicted variable value;
and the user retention cause and effect determination submodule is used for determining the cause and effect of the intervention variable on the user retention according to the first cause and effect or the second cause and effect of each account.
In one embodiment, the predictor variable value determination sub-module may further include the following units:
the sample set organization unit is used for organizing all accounts which receive intervention into an experiment sample set and organizing all accounts which do not receive intervention into a control sample set;
the model training unit is used for training a first machine model according to the experimental sample set and training a second machine model according to the control sample set;
the model prediction unit is used for predicting a first prediction variable value of each account in the experiment sample set under the condition of not accepting intervention by adopting the second machine model; and predicting a second predicted variable value of each account in the control sample set under the condition of receiving the intervention by adopting the first machine model.
In one embodiment, the model training unit is specifically configured to:
selecting a target covariate from covariates according to the intervention variable, wherein the covariate comprises: the other specified characteristic variables except the specified characteristic variable corresponding to the intervention variable in the specified characteristic variables;
and training a first machine model by adopting a preset machine learning algorithm based on the variable values corresponding to the target covariates of the accounts in the experiment sample set and the retention information of the accounts.
In one embodiment, the first causal effect determination submodule is specifically configured to:
calculating a difference between the first observed variable value and the first predicted variable value as the first causal effect.
In one embodiment, the user saved causal effect determination submodule may further include:
the first average causal effect calculation unit is used for calculating the average value of all the first causal effects as the first average causal effect of the experimental sample set;
a second average causal effect calculation unit for calculating an average of all second causal effects as a second average causal effect of the set of control samples;
a tendency value determination unit for determining a tendency value reflecting the probability of the user accepting intervention given the target covariate;
and the weighting calculation unit is used for performing weighting calculation on the first average causal effect and the second average causal effect by taking the tendency value as a weight to obtain the causal effect of the intervention variable on the user retention under the given target covariate.
In one embodiment, the tendency value determination unit is specifically configured to:
and inputting the variable value of the target covariate into a trained probability model, and outputting the probability of the user to be intervened predicted under the target covariate as a tendency value by the probability model.
In one embodiment, if there are more than one covariate of interest, the user retained causal effect determination submodule is further configured to:
and calculating the average expectation of the causal effect corresponding to each target covariate, and obtaining the causal effect of the intervention variable on the user retention.
In one embodiment, when there are a plurality of the intervention variables, the apparatus may further include:
and the key factor determining module is used for selecting the intervention variable with the largest causal effect as the key factor influencing the retention of the user according to the causal effect of each intervention variable on the retention of the user.
It should be noted that the above feature evaluation apparatus that affects user retention provided in the embodiment of the present application can execute the method provided in the first embodiment of the present application, and has corresponding functional modules and beneficial effects for executing the method.
EXAMPLE III
Fig. 4 is a schematic structural diagram of an electronic device according to a third embodiment of the present application, as shown in fig. 4, the electronic device includes a processor 410, a memory 420, an input device 430, and an output device 440; the number of the processors 410 in the electronic device may be one or more, and one processor 410 is taken as an example in fig. 4; the processor 410, the memory 420, the input device 430 and the output device 440 in the electronic apparatus may be connected by a bus or other means, and the bus connection is exemplified in fig. 4.
The memory 420 serves as a computer-readable storage medium for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the methods in the embodiments of the present application. The processor 410 executes various functional applications of the electronic device and data processing by executing software programs, instructions and modules stored in the memory 420, that is, implements the above-described method.
The memory 420 may mainly include a program storage area and a data storage area, wherein the program storage area
The operating system and the application program required by at least one function can be stored; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 420 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 420 may further include memory located remotely from processor 410, which may be connected to an electronic device over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 430 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the electronic apparatus. The output device 440 may include a display device such as a display screen.
Example four
The fourth embodiment of the present application further provides a storage medium containing computer-executable instructions, which when executed by a processor of a server, are configured to perform the method of any one of the first embodiment.
From the above description of the embodiments, it is obvious for those skilled in the art that the present application can be implemented by software and necessary general hardware, and certainly can be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods described in the embodiments of the present application.
It should be noted that, in the embodiment of the apparatus, the included units and modules are merely divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only used for distinguishing one functional unit from another, and are not used for limiting the protection scope of the application.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present application and the technical principles employed. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, although the present application has been described in more detail with reference to the above embodiments, the present application is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present application, and the scope of the present application is determined by the scope of the appended claims.
Claims (13)
1. A method of feature assessment that affects user retention, the method comprising:
acquiring corresponding specified characteristic values of each account in the current platform according to the specified characteristic variables;
acquiring retention information of each account, wherein the retention information is used for reflecting whether the current account is a retention account;
determining an intervention variable according to the acquired specified characteristic value;
and determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information.
2. The method of claim 1, wherein determining intervention variables from the collected specified feature values comprises:
determining a target specified characteristic variable from the plurality of specified characteristic variables;
and determining a target specified characteristic value from one or more specified characteristic values corresponding to the target specified characteristic variable to generate an intervention variable.
3. The method of claim 1 or 2, wherein said determining a causal effect of the intervention variable on user retention using a causal inference algorithm in conjunction with the retention information comprises:
acquiring a first observation variable value corresponding to the account which receives the intervention and a second observation variable value corresponding to the account which does not receive the intervention according to the intervention variable;
determining a first predictive variable value of the account which receives the intervention under the condition of not receiving the intervention, and determining a second predictive variable value of the account which does not receive the intervention under the condition of receiving the intervention;
determining a first causal effect of the account undergoing intervention based on the first observed variable value and the first predicted variable value;
determining a second causal effect of the account not accepting intervention based on the second observed variable value and the second predicted variable value;
and determining the causal effect of the intervention variable on the user retention according to the first causal effect or the second causal effect of each account.
4. The method of claim 3, wherein determining a first predictive variable value for the account undergoing intervention without undergoing intervention and determining a second predictive variable value for the account not undergoing intervention comprises:
organizing all accounts which receive intervention into an experimental sample set, and organizing all accounts which do not receive intervention into a control sample set;
training a first machine model from the set of experimental samples, and training a second machine model from the set of control samples;
predicting a first prediction variable value of each account in the experiment sample set under the condition of not accepting intervention by adopting the second machine model;
and predicting a second prediction variable value of each account in the control sample set under the condition of receiving the intervention by adopting the first machine model.
5. The method of claim 4, wherein training a first machine model from the set of experimental samples comprises:
selecting a target covariate from covariates according to the intervention variable, wherein the covariate comprises: the other specified characteristic variables except the specified characteristic variable corresponding to the intervention variable in the specified characteristic variables;
and training a first machine model by adopting a preset machine learning algorithm based on the variable values corresponding to the target covariates of the accounts in the experiment sample set and the retention information of the accounts.
6. The method of claim 5, wherein said determining a first causal effect of said account undergoing intervention based on said first observed variable value and said first predicted variable value comprises:
calculating a difference between the first observed variable value and the first predicted variable value as the first causal effect.
7. The method according to claim 5 or 6, wherein the determining the causal effect of the intervention variable on the user retention according to the first causal effect or the second causal effect of each account comprises:
calculating an average of all first causal effects as a first average causal effect for the set of experimental samples;
calculating an average of all second causal effects as a second average causal effect for the set of control samples;
determining a trend value reflecting a probability of a user accepting intervention given a target covariate;
and taking the tendency value as a weight, and carrying out weighted calculation on the first average causal effect and the second average causal effect to obtain the causal effect of the intervention variable on the user retention under the given target covariate.
8. The method of claim 7, wherein determining the trend value comprises:
and inputting the variable value of the target covariate into a trained probability model, and outputting the probability of the user to be intervened predicted under the target covariate as a tendency value by the probability model.
9. The method of claim 7, wherein if there are a plurality of covariates, the determining the causal effect of the intervention variable on the user retention based on the first causal effect or the second causal effect for each account further comprises:
and calculating the average expectation of the causal effect corresponding to each target covariate, and obtaining the causal effect of the intervention variable on the user retention.
10. The method according to claim 1 or 2, wherein when there are a plurality of said intervention variables, after said determining a causal effect of said intervention variables on user retention using a causal inference algorithm in combination with said retention information, the method further comprises:
and selecting the intervention variable with the largest causal effect as a key factor influencing the retention of the user according to the causal effect of each intervention variable on the retention of the user.
11. An apparatus for feature assessment that affects user retention, the apparatus comprising:
the system comprises a designated characteristic value acquisition module, a characteristic value analysis module and a characteristic value analysis module, wherein the designated characteristic value acquisition module is used for acquiring corresponding designated characteristic values of accounts in a current platform according to a plurality of designated characteristic variables;
the retention information acquisition module is used for acquiring retention information of each account, and the retention information is used for reflecting whether the current account is a retention account;
the intervention variable determining module is used for determining an intervention variable according to the acquired specified characteristic value;
and the causal effect determination module is used for determining the causal effect of the intervention variable on the user retention by adopting a causal inference algorithm in combination with the retention information.
12. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1-10 when executing the program.
13. 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 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011613766.9A CN112685674A (en) | 2020-12-30 | 2020-12-30 | Feature evaluation method and device influencing user retention |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011613766.9A CN112685674A (en) | 2020-12-30 | 2020-12-30 | Feature evaluation method and device influencing user retention |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112685674A true CN112685674A (en) | 2021-04-20 |
Family
ID=75455367
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011613766.9A Pending CN112685674A (en) | 2020-12-30 | 2020-12-30 | Feature evaluation method and device influencing user retention |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112685674A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113112311A (en) * | 2021-05-12 | 2021-07-13 | 北京百度网讯科技有限公司 | Method for training causal inference model, information prompting method and device |
| CN114186096A (en) * | 2021-12-10 | 2022-03-15 | 北京达佳互联信息技术有限公司 | Information processing method and device |
| CN115794175A (en) * | 2023-02-06 | 2023-03-14 | 广东省科学技术情报研究所 | Big data-based technology research and development evaluation system and method |
Citations (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100333134A1 (en) * | 2009-06-30 | 2010-12-30 | Mudd Advertising | System, method and computer program product for advertising |
| US20130218814A1 (en) * | 2012-02-20 | 2013-08-22 | Xerox Corporation | Method and system for the dynamic allocation of resources based on fairness, throughput, and user behavior measurement |
| CN105005909A (en) * | 2015-06-17 | 2015-10-28 | 深圳市腾讯计算机系统有限公司 | Method and device for predicting lost users |
| CN106294658A (en) * | 2016-08-04 | 2017-01-04 | 腾讯科技（深圳）有限公司 | The quick methods of exhibiting of webpage and device |
| CN106529727A (en) * | 2016-11-18 | 2017-03-22 | 腾讯科技（深圳）有限公司 | User loss prediction model generation method and related device |
| CN106959916A (en) * | 2017-02-28 | 2017-07-18 | 腾讯科技（深圳）有限公司 | Client retains influence detection method and device |
| CN107230091A (en) * | 2016-03-23 | 2017-10-03 | 滴滴（中国）科技有限公司 | Order matching process and device are asked in share-car |
| CN107292150A (en) * | 2016-04-13 | 2017-10-24 | 平安科技（深圳）有限公司 | Save user identification confirmation method and apparatus in information processing from damage |
| CN107437095A (en) * | 2017-07-24 | 2017-12-05 | 腾讯科技（深圳）有限公司 | Classification determines method and device |
| CN108053322A (en) * | 2017-12-15 | 2018-05-18 | 东峡大通（北京）管理咨询有限公司 | The customer investment return evaluation method and system of vehicle |
| CN108170909A (en) * | 2017-12-13 | 2018-06-15 | 中国平安财产保险股份有限公司 | Model output method, equipment and the storage medium of a kind of intelligent modeling |
| CN108364195A (en) * | 2018-02-09 | 2018-08-03 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN108415981A (en) * | 2018-02-09 | 2018-08-17 | 平安科技（深圳）有限公司 | Data dimension generation method, device, equipment and computer readable storage medium |
| CN108537587A (en) * | 2018-04-03 | 2018-09-14 | 广州优视网络科技有限公司 | It is lost in user's method for early warning, device, computer readable storage medium and server |
| CN108648000A (en) * | 2018-04-24 | 2018-10-12 | 腾讯科技（深圳）有限公司 | Method and device, the electronic equipment that life cycle is assessed are retained to user |
| CN109492891A (en) * | 2018-10-26 | 2019-03-19 | 阿里巴巴集团控股有限公司 | Customer churn prediction technique and device |
| CN109858970A (en) * | 2019-02-02 | 2019-06-07 | 中国银行股份有限公司 | A kind of user's behavior prediction method, apparatus and storage medium |
| CN110585726A (en) * | 2019-09-16 | 2019-12-20 | 腾讯科技（深圳）有限公司 | User recall method, device, server and computer readable storage medium |
| CN111260382A (en) * | 2018-11-30 | 2020-06-09 | 北京嘀嘀无限科技发展有限公司 | Prediction processing method and device for loss probability |
| CN111311030A (en) * | 2020-03-27 | 2020-06-19 | 中国工商银行股份有限公司 | User credit risk prediction method and device based on influence factor detection |
| CN111723973A (en) * | 2020-05-15 | 2020-09-29 | 西安交通大学 | Learning effect optimization method based on user behavior causal relationship in MOOC log data |
| CN111988634A (en) * | 2020-08-14 | 2020-11-24 | 广州市百果园信息技术有限公司 | Anchor selection method and device, computer readable storage medium and electronic equipment |
| CN112035549A (en) * | 2020-08-31 | 2020-12-04 | 中国平安人寿保险股份有限公司 | Data mining method and device, computer equipment and storage medium |
| CN112070533A (en) * | 2020-08-28 | 2020-12-11 | 上海连尚网络科技有限公司 | Method and equipment for predicting user retention |
-
2020
- 2020-12-30 CN CN202011613766.9A patent/CN112685674A/en active Pending
Patent Citations (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100333134A1 (en) * | 2009-06-30 | 2010-12-30 | Mudd Advertising | System, method and computer program product for advertising |
| US20130218814A1 (en) * | 2012-02-20 | 2013-08-22 | Xerox Corporation | Method and system for the dynamic allocation of resources based on fairness, throughput, and user behavior measurement |
| CN105005909A (en) * | 2015-06-17 | 2015-10-28 | 深圳市腾讯计算机系统有限公司 | Method and device for predicting lost users |
| CN107230091A (en) * | 2016-03-23 | 2017-10-03 | 滴滴（中国）科技有限公司 | Order matching process and device are asked in share-car |
| CN107292150A (en) * | 2016-04-13 | 2017-10-24 | 平安科技（深圳）有限公司 | Save user identification confirmation method and apparatus in information processing from damage |
| CN106294658A (en) * | 2016-08-04 | 2017-01-04 | 腾讯科技（深圳）有限公司 | The quick methods of exhibiting of webpage and device |
| CN106529727A (en) * | 2016-11-18 | 2017-03-22 | 腾讯科技（深圳）有限公司 | User loss prediction model generation method and related device |
| CN106959916A (en) * | 2017-02-28 | 2017-07-18 | 腾讯科技（深圳）有限公司 | Client retains influence detection method and device |
| CN107437095A (en) * | 2017-07-24 | 2017-12-05 | 腾讯科技（深圳）有限公司 | Classification determines method and device |
| CN108170909A (en) * | 2017-12-13 | 2018-06-15 | 中国平安财产保险股份有限公司 | Model output method, equipment and the storage medium of a kind of intelligent modeling |
| CN108053322A (en) * | 2017-12-15 | 2018-05-18 | 东峡大通（北京）管理咨询有限公司 | The customer investment return evaluation method and system of vehicle |
| CN108415981A (en) * | 2018-02-09 | 2018-08-17 | 平安科技（深圳）有限公司 | Data dimension generation method, device, equipment and computer readable storage medium |
| CN108364195A (en) * | 2018-02-09 | 2018-08-03 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN108537587A (en) * | 2018-04-03 | 2018-09-14 | 广州优视网络科技有限公司 | It is lost in user's method for early warning, device, computer readable storage medium and server |
| CN108648000A (en) * | 2018-04-24 | 2018-10-12 | 腾讯科技（深圳）有限公司 | Method and device, the electronic equipment that life cycle is assessed are retained to user |
| CN109492891A (en) * | 2018-10-26 | 2019-03-19 | 阿里巴巴集团控股有限公司 | Customer churn prediction technique and device |
| CN111260382A (en) * | 2018-11-30 | 2020-06-09 | 北京嘀嘀无限科技发展有限公司 | Prediction processing method and device for loss probability |
| CN109858970A (en) * | 2019-02-02 | 2019-06-07 | 中国银行股份有限公司 | A kind of user's behavior prediction method, apparatus and storage medium |
| CN110585726A (en) * | 2019-09-16 | 2019-12-20 | 腾讯科技（深圳）有限公司 | User recall method, device, server and computer readable storage medium |
| CN111311030A (en) * | 2020-03-27 | 2020-06-19 | 中国工商银行股份有限公司 | User credit risk prediction method and device based on influence factor detection |
| CN111723973A (en) * | 2020-05-15 | 2020-09-29 | 西安交通大学 | Learning effect optimization method based on user behavior causal relationship in MOOC log data |
| CN111988634A (en) * | 2020-08-14 | 2020-11-24 | 广州市百果园信息技术有限公司 | Anchor selection method and device, computer readable storage medium and electronic equipment |
| CN112070533A (en) * | 2020-08-28 | 2020-12-11 | 上海连尚网络科技有限公司 | Method and equipment for predicting user retention |
| CN112035549A (en) * | 2020-08-31 | 2020-12-04 | 中国平安人寿保险股份有限公司 | Data mining method and device, computer equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 董伟;李建红;陶金虎;: "在线健康社区活跃用户识别及其交互类型分析", 文献与数据学报, no. 01, 30 March 2020 (2020-03-30), pages 91 - 103 * |
| 魏玲;郭新悦;: "基于改进RFM与GMDH算法的MOOC用户流失预测", 中国远程教育, no. 09, 9 September 2020 (2020-09-09), pages 43 - 47 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113112311A (en) * | 2021-05-12 | 2021-07-13 | 北京百度网讯科技有限公司 | Method for training causal inference model, information prompting method and device |
| CN113112311B (en) * | 2021-05-12 | 2023-07-25 | 北京百度网讯科技有限公司 | Method for training causal inference model and information prompting method and device |
| CN114186096A (en) * | 2021-12-10 | 2022-03-15 | 北京达佳互联信息技术有限公司 | Information processing method and device |
| CN115794175A (en) * | 2023-02-06 | 2023-03-14 | 广东省科学技术情报研究所 | Big data-based technology research and development evaluation system and method |
| CN115794175B (en) * | 2023-02-06 | 2023-05-02 | 广东省科学技术情报研究所 | Technology research and development evaluation system and method based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112685674A (en) | Feature evaluation method and device influencing user retention | |
| CN111667010A (en) | Sample evaluation method, device and equipment based on artificial intelligence and storage medium | |
| CN111881023B (en) | Software aging prediction method and device based on multi-model comparison | |
| CN111290967A (en) | Software defect prediction method, device, equipment and storage medium | |
| CN112200392B (en) | Service prediction method and device | |
| CN111160959B (en) | User click conversion prediction method and device | |
| CN113240510B (en) | Abnormal user prediction method, device, equipment and storage medium | |
| CN112671985A (en) | Agent quality inspection method, device, equipment and storage medium based on deep learning | |
| CN112508695A (en) | Financial risk prediction method and device based on market risk and electronic equipment | |
| CN112966865A (en) | Number portability prediction method, device and equipment | |
| CN116186221A (en) | Big data analysis method and system applied to online dialogue platform | |
| CN110910241B (en) | Cash flow evaluation method, apparatus, server device and storage medium | |
| CN110704614B (en) | Information processing method and device for predicting user group type in application | |
| CN110852322B (en) | Method and device for determining region of interest | |
| CN109992778B (en) | Resume document distinguishing method and device based on machine learning | |
| CN111599382A (en) | Voice analysis method, device, computer equipment and storage medium | |
| CN115660101A (en) | Data service providing method and device based on service node information | |
| CN113934863B (en) | Food safety risk prediction method and device, electronic equipment and medium | |
| CN116186603A (en) | Abnormal user identification method and device, computer storage medium and electronic equipment | |
| CN110334353B (en) | Analysis method, device, equipment and storage medium for word sequence recognition performance | |
| CN114863994A (en) | Pollution assessment method, device, electronic equipment and storage medium | |
| CN114420219A (en) | Construction method, prediction method and device of relative retention time prediction model | |
| CN110852854B (en) | Method for generating quantitative gain model and method for evaluating risk control strategy | |
| CN113095589A (en) | Population attribute determination method, device, equipment and storage medium | |
| CN116208513A (en) | Gateway health degree prediction method and device |
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 |