CN113205188A - Method and device for determining target account, storage medium and electronic device - Google Patents
Method and device for determining target account, storage medium and electronic device Download PDFInfo
- Publication number
- CN113205188A CN113205188A CN202110495273.8A CN202110495273A CN113205188A CN 113205188 A CN113205188 A CN 113205188A CN 202110495273 A CN202110495273 A CN 202110495273A CN 113205188 A CN113205188 A CN 113205188A
- Authority
- CN
- China
- Prior art keywords
- account
- probability
- target
- event
- prediction model
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a method, a device, a storage medium and an electronic device for determining a target account, wherein the method comprises the following steps: predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set marked with a trigger tag, and the trigger tag is used for indicating whether the account sample in the account sample set triggers the target event or not; determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history corresponding to the target event of each account; and determining a target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account. The method and the device solve the technical problem that the credibility of the account which is determined from the plurality of accounts and has the incidence relation with the event is low.
Description
Technical Field
The present application relates to the field of data processing, and in particular, to a method and an apparatus for determining a target account, a storage medium, and an electronic apparatus.
Background
With the increasing development of network life, people move and operate more and more frequently on the network, and network data is increased rapidly. For the field that needs to analyze the personal characteristics to predict the personal behavior, for example, the residential characteristics of a certain area are analyzed to predict the residential demand, and the analysis can be performed according to the network data generated by residents on the network. However, the current way of predicting personal behavior by combining network data is generally to evaluate the collected personal network data through the experience of domain experts so as to analyze personal behavior.
On one hand, the prediction mode may depend too much on the knowledge framework of the domain expert, so that more subjective factors of the domain expert may be introduced in the evaluation process, for example, the domain experts with different experiences give different evaluation criteria, resulting in lower reliability of the evaluation result. On the other hand, the evaluation result may have a larger limitation due to a single field related by the field expert, and the reliability of the evaluation result is further influenced.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The application provides a method and a device for determining a target account, a storage medium and an electronic device, which are used for solving the technical problem that the credibility of an account which is determined from a plurality of accounts and has an incidence relation with an event is low in the related art.
According to an aspect of an embodiment of the present application, there is provided a method for determining a target account, including: predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set marked with a trigger tag, and the trigger tag is used for indicating whether the target event is triggered by the account sample in the account sample set; determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history of each account corresponding to the target event; and determining a target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
Optionally, determining, according to the first probability corresponding to each account and the second probability corresponding to each account, that the third probability corresponding to each account includes one of: determining a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account; determining a geometric mean of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account.
Optionally, determining a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account includes: acquiring a first weight value corresponding to the event prediction model and a second weight value corresponding to an analysis process of account use history, wherein the first weight value is used for indicating the reliability of the event prediction model, and the second weight value is used for indicating the reliability of the analysis process of account use history; and summing and calculating the first weight value as the weight of the first probability of each account and the second weight value as the weight of the second probability of each account to obtain the third probability corresponding to each account.
Optionally, before determining the third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, the method further includes: acquiring account data generated in the using process of each account; extracting a characteristic value of a target attribute characteristic from the account data, wherein the target attribute characteristic is an attribute characteristic corresponding to the target event in all attribute characteristics of the account data; analyzing the characteristic value of the target attribute characteristic of each account to obtain the second probability corresponding to each account.
Optionally, predicting, by the event prediction model, the first probability that the each account of the plurality of accounts triggers the target event comprises: acquiring a characteristic value of a target attribute characteristic included in each account, wherein the target attribute characteristic is an attribute characteristic corresponding to the target event in all attribute characteristics of the account data; inputting a characteristic value of the target attribute feature included in each account into the event prediction model; and acquiring a target probability value output by the event prediction model as the first probability corresponding to each account.
Optionally, before inputting the feature value of the target attribute feature included in each account into the event prediction model, the method further includes: inputting the account sample set into the initial event prediction model; adjusting model parameters of the initial event prediction model according to a comparison value between an initial probability value output by the initial event prediction model and the trigger tag until the comparison value between the initial probability value and the trigger tag falls into a preset threshold range; determining a model parameter which enables a comparison value between the initial probability value and the trigger tag to fall within a preset threshold range as a target model parameter of the event prediction model.
Optionally, according to the third probability corresponding to each account, determining, from the multiple accounts, a target account having an association relationship with the target event includes one of: acquiring the account with the target number with the highest third probability from the plurality of accounts as the target account; and acquiring the account with the third probability higher than the probability threshold from the plurality of accounts as the target account.
According to another aspect of the embodiments of the present application, there is also provided an apparatus for determining a target account, including: the prediction module predicts a first probability that each account in a plurality of accounts triggers a target event through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model through an account sample set marked with a trigger label, and the trigger label is used for indicating whether the account sample in the account sample set triggers the target event; a first determining module, configured to determine a third probability corresponding to each account according to the first probability corresponding to each account and a second probability corresponding to each account, where the second probability corresponding to each account is obtained by analyzing a usage history of each account corresponding to the target event; and the second determining module is used for determining a target account which has an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
According to another aspect of the embodiments of the present application, there is also provided a storage medium including a stored program which, when executed, performs the above-described method.
According to another aspect of the embodiments of the present application, there is also provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the above method through the computer program.
The method can be applied to the technical field of marketing intelligence for prediction and optimization, in the embodiment of the application, a first probability that each account in a plurality of accounts triggers a target event is predicted through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model through an account sample set marked with a trigger label, and the trigger label is used for indicating whether the account sample in the account sample set triggers the target event or not; determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history corresponding to the target event of each account; and determining a mode of the target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
In the process of determining a target account from a plurality of accounts, predicting a first probability corresponding to each account in the plurality of accounts through an event prediction model, wherein the model is obtained by training an initial event prediction model through an account sample set marked with a trigger label for indicating whether an account sample in the account sample set triggers a target event or not, so that the first probability can objectively and accurately represent the association between the account and the target event, determining a third probability corresponding to each account according to a second probability obtained by analyzing the use history of each account to represent the final probability of triggering the target event by each account, and since the third probability is obtained by combining the prediction and analysis processes of a plurality of dimensions, the third probability can more accurately represent the possibility of triggering the target event by the account, so that the target account which is determined according to the third probability and has an association relation with the target event is more credible, therefore, the technical effect of improving the reliability of the account which is determined from the plurality of accounts and has the incidence relation with the event is achieved, and the technical problem that the reliability of the account which is determined from the plurality of accounts and has the incidence relation with the event is low is solved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
FIG. 1 is a diagram illustrating a hardware environment for a method for determining a target account according to an embodiment of the present application;
FIG. 2 is a flowchart of an alternative method for determining a target account according to an embodiment of the present application;
FIG. 3 is a flow chart of an alternative account ranking according to an embodiment of the present application;
FIG. 4 is a block diagram of an alternative scoring ranking according to an embodiment of the present application;
FIG. 5 is a schematic diagram of an alternative target account determination apparatus according to an embodiment of the present application;
fig. 6 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an aspect of the embodiments of the present application, an embodiment of a method for determining a target account is provided.
Alternatively, in this embodiment, the method for determining the target account may be applied to a hardware environment formed by the terminal 101 and the server 103 as shown in fig. 1. As shown in fig. 1, a server 103 is connected to a terminal 101 through a network, which may be used to provide services (such as data computing services) for the terminal or a client installed on the terminal, and a database may be provided on the server or separately from the server, and is used to provide data storage services for the server 103, and the network includes but is not limited to: the terminal 101 is not limited to a PC, a mobile phone, a tablet computer, and the like. The method for determining the target account number in the embodiment of the present application may be executed by the server 103, the terminal 101, or both the server 103 and the terminal 101. The method for determining the target account performed by the terminal 101 according to the embodiment of the present application may also be performed by a client installed thereon.
Fig. 2 is a flowchart of an optional target account determination method according to an embodiment of the present application, and as shown in fig. 2, the method may include the following steps:
step S202, predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set labeled with a trigger label, and the trigger label is used for indicating whether the target event is triggered by the account sample in the account sample set;
step S204, determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history of each account;
step S206, determining a target account from the plurality of accounts according to the third probability corresponding to each account.
Through the above steps S202 to S206, in the process of determining the target account from the multiple accounts, the first probability corresponding to each account in the multiple accounts is predicted through an event prediction model, which is obtained by training an initial event prediction model using an account sample set labeled with a trigger tag for indicating whether an account sample in the account sample set triggers the target event, so that the first probability can more objectively and accurately represent the association between the account and the target event, and then the third probability corresponding to each account is determined through the second probability obtained by analyzing the usage history of each account to represent the final probability that each account triggers the target event, because the third probability is obtained through the prediction and analysis processes of multiple dimensions, the third probability can more accurately represent the possibility that the account triggers the target event, the target account which is determined according to the third probability and has the incidence relation with the target event is more credible, so that the technical effect of improving the credibility of the account which is determined from the multiple accounts and has the incidence relation with the event is achieved, and the technical problem that the credibility of the account which is determined from the multiple accounts and has the incidence relation with the event is lower is solved.
In the technical solution provided in step S202, the target event may include, but is not limited to, that the target account purchases a certain product, a certain preset rule is violated, and the like, and the solution is not limited thereto.
Optionally, in this embodiment, the event prediction model may be implemented by, but not limited to, using a classification model, such as: LR (Logistic Regression) model, XGBoost model, deep fm model, and the like.
In the technical solution provided in step S204, the method for determining the third probability corresponding to each account may be to calculate the first probability of each account and the second probability of each account to obtain the third probability, or to select one of the first probability and the second probability as the third probability by comparing the first probability and the second probability.
Optionally, in this embodiment, the usage history corresponding to the target event for each account may include, but is not limited to, browsing data, operation data, interaction data, and the like corresponding to the target event generated during the usage of the account.
Optionally, in this embodiment, parsing the usage history of each account may include, but is not limited to, processing each account usage data using an algorithm or a parsing model, such as: and normalizing the use data, the operation data and the interaction data of other accounts, and calculating the processed data by using a formula or an algorithm model so as to obtain a second probability corresponding to the account.
In the technical solution provided in step S206, the target account may be one or more of the multiple accounts with the largest or smallest third probability, or an account with a third probability greater than a certain threshold, or an account with a third probability located in a certain threshold interval, which is not limited in this solution.
Optionally, in this embodiment, after the target account is determined in the multiple accounts, a notification message may be generated according to the determined target account and the third probability of the target account, and the notification message is sent to the target account, a certain preset account, or a certain preset terminal device, or a corresponding operation instruction may be generated according to the determined target account and the third probability of the target account, where the operation instruction is used to instruct to perform some preset operations on the target account, for example, to limit some functions of the target account, to open some functions of the target account, to recommend some items to the target account, and the like, which is not limited in this scheme.
As an optional embodiment, determining the third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account respectively includes one of:
s11, determining a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account;
s12, determining a geometric mean of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account.
Optionally, in this embodiment, the method of calculating the geometric mean of the first probability corresponding to each account and the second probability of each account may be to calculate a quadratic root of a product of the first probability and the second probability of each account.
Through the steps, the first probability and the second probability of each account are subjected to weighted summation calculation, or the geometric mean value of the first probability and the second probability of each account is calculated, so that the first probability predicted by the prediction model and the second probability obtained by account checking analysis are fused, and the accuracy of the obtained third probability of each account is improved.
As an optional embodiment, determining a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account includes:
s21, acquiring a first weight value corresponding to the event prediction model and a second weight value corresponding to an analysis process of account use history, wherein the first weight value is used for indicating the credibility of the event prediction model, and the second weight value is used for indicating the credibility of the analysis process of account use history;
s22, summing the first weight value as the weight of the first probability of each account and the second weight value as the weight of the second probability of each account, so as to obtain the third probability corresponding to each account.
Optionally, in this embodiment, both the first weight value and the second weight value may be preset fixed values, or an average value of the first weight value and an average value of the second weight value of each of the multiple reference accounts may be respectively determined as the first weight value and the second weight value of each account, or the first weight value and the second weight value obtained by the last calculation may be obtained by trimming according to a ratio of the first probability and the second probability of the account, which is not limited in this embodiment.
Through the steps, the credibility of the event prediction model and the credibility of the analysis process of the account use history are obtained, the first weight value corresponding to the event prediction model and the second weight value corresponding to the analysis process of the account use history are obtained, the first probability output by the event prediction model is weighted by using the first weight value, the second probability obtained by the analysis of the account use history is weighted by using the second weight value, the summation calculation is carried out, and the third probability of the account is obtained, so that the third probability of the account is influenced by the credibility of the event prediction model and the credibility of the analysis process of the account use history, and the accuracy of the obtained third probability of the account is further ensured.
As an optional embodiment, before determining the third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, the method further includes:
s31, acquiring the account data generated in the using process of each account;
s32, extracting a feature value of a target attribute feature from the account data, where the target attribute feature is an attribute feature corresponding to the target event in all attribute features of the account data;
and S33, analyzing the characteristic value of the target attribute characteristic of each account to obtain the second probability corresponding to each account.
Optionally, in this embodiment, the acquiring of the account data may be implemented by acquiring a log file of the account, and performing screening, extraction, and the like on the log file, or may be implemented by performing real-time monitoring on the account.
Optionally, in this embodiment, the account data may include, but is not limited to, attribute feature data of various attributes of the account, browsing data of the account, operation data of the account, interaction data with other accounts, and the like.
Optionally, in this embodiment, the feature value of the target attribute feature extracted from the account data may be a feature value of the target attribute feature directly extracted from the account data, or may be a feature value obtained by extracting data of the target attribute feature from the account data and calculating the target attribute feature, or a feature value corresponding to the target attribute feature and matched with a preset feature value, which is not limited in this embodiment.
Optionally, in this embodiment, the type of the target attribute feature extracted from the account data may be preset, for example, 1, 2, 3, 5, and so on number of target attribute features may be set to be extracted from the account data.
Optionally, in this embodiment, in the process of analyzing the feature values of the target attribute features, weights may be set for the feature values, different weights of the feature values are different, the weights of each target attribute feature are added, and the sum of the weights is used as the second probability; the ratio between the feature values may also be calculated, different weights may be set for different target attribute features, the weight value of the target attribute feature is used as the weight of the corresponding feature value, and the summation calculation is performed, and the calculation result is used as a second probability, for example: extracting four characteristics corresponding to a target event in account data of a plurality of accounts, wherein the four characteristics are respectively a school calendar, a graduation school, a month income and deposit information, presetting attribute categories of the characteristics as education and property categories, presetting attribute categories of the characteristics that the school calendar and the graduation school belong to the education category, the month income and the deposit information belong to the property category, presetting a weight value of the characteristics of the education category as 0.4, presetting a weight value of the characteristics of the property category as 0.6, presetting each characteristic score group, for example, uniformly setting 4 score groups in a range of 0 to 100, wherein the 4 score groups are respectively [0,25 ], [25,50 ], [50,75 and [ 100], setting the weight values of each score group as 0.1, 0.2,0.3 and 0.4, normalizing the four attributes and calculating to obtain characteristic values corresponding to the characteristics of the school calendar as 75, 0, 0.2,0.3 and 0.4, Since the feature value of the feature of graduation school is 45 points, the feature value of the feature of monthly income is 88 points, and the feature value of deposit information is 60 points, it is determined that the weight value of the feature of academic calendar is 0.4, the weight value of the feature of graduation school is 0.2, the weight value of the feature of monthly income is 0.4, and the weight value of the feature of deposit information is 0.3, and the second probability of the account is [ (0.4+0.2) × 0.4+ (0.4+0.3) × 0.6 ]% 100% 66% by summing up the weight values of the features and the attribute categories to which the features belong, and the second probability of the account is 66% can be obtained.
As an alternative embodiment, predicting the first probability that the each account of the plurality of accounts triggers the target event by the event prediction model comprises:
s41, acquiring a feature value of a target attribute feature included in each account, where the target attribute feature is an attribute feature corresponding to the target event in all attribute features of the account data;
s42, inputting the characteristic value of the target attribute feature included in each account into the event prediction model;
and S43, acquiring a target probability value output by the event prediction model as the first probability corresponding to each account.
Optionally, in this embodiment, the target attribute features are the same as the number and the types of the target attribute features used for calculating the second probability.
Optionally, in this embodiment, the event prediction model may be a convolutional neural network model, and the feature value of the target attribute feature included in each account is input into a target input layer of the event prediction model, where the event prediction model includes a target input layer, a target convolutional layer, and a target fully-connected layer, which are connected in sequence, the target input layer is configured to receive an input parameter of the event prediction model, the target convolutional layer is configured to extract a parameter feature from the input parameter, the target fully-connected layer is configured to predict a probability that an account with the parameter feature triggers a target event, and a target probability value output by the target fully-connected layer is obtained as a first probability corresponding to each account.
As an optional embodiment, before inputting the feature value of the target attribute feature included in each account into the event prediction model, the method further includes:
s51, inputting the account sample set into the initial event prediction model;
s52, adjusting the model parameters of the initial event prediction model according to the comparison value between the initial probability value output by the initial event prediction model and the trigger tag until the comparison value between the initial probability value and the trigger tag falls into a preset threshold range;
and S53, determining the model parameters which enable the comparison value between the initial probability value and the trigger label to fall within a preset threshold range as the target model parameters of the event prediction model.
Optionally, in this embodiment, when the event prediction model is a convolutional neural network model, the process of training the model may be to input the account sample set into an initial input layer included in the initial event prediction model, where the initial event prediction model includes an initial input layer, an initial convolutional layer, and an initial fully-connected layer that are connected in sequence; adjusting model parameters of the initial input layer, the initial convolutional layer and the initial full-connection layer according to a comparison value between an initial probability value output by the initial full-connection layer and the trigger label until the comparison value between the initial probability value and the trigger label falls into a preset threshold range; and determining the model parameters of the comparison value between the initial probability value and the trigger label in a preset threshold range as the model parameters of the target input layer, the target convolution layer and the target full-connection layer to obtain the event prediction model.
Optionally, in this embodiment, when the event prediction model is a logistic regression model, the training model may be an account sample set input value algorithm model labeled with a trigger tag, and the model parameters of the initial event prediction model are adjusted according to a comparison value between an initial probability value output by the initial event prediction model and the trigger tag.
As an optional embodiment, according to the third probability corresponding to each account, determining a target account from the plurality of accounts includes one of:
s61, acquiring the account with the target number with the highest third probability from the plurality of accounts as the target account;
and S62, acquiring an account with the third probability higher than the probability threshold from the plurality of accounts as the target account.
Optionally, in this embodiment, the target number may be flexibly set according to requirements, for example, the target number may be set to be 1, 2, 5, and the like.
Optionally, the step of obtaining the account with the target number with the highest third probability from the multiple accounts may be to sort the accounts according to the third probability, and extract the account with the target number from front to back as the target account; and sorting the accounts according to the third probability, and extracting the account with the first rank and the account with the first rank in parallel as target accounts.
Fig. 3 is a flow chart of an alternative account sorting according to an embodiment of the present application, as shown in fig. 3:
s301, acquiring the characteristics of each account in a plurality of accounts as original data, and constructing an original data table X, wherein the original data table comprises M rows and N columns, each row represents an account to be sorted, each column represents one characteristic of the account, and each account comprises N characteristics.
S302, K characteristics of each account are screened out from an original data table X, the screened K characteristics are characteristics which have corresponding relations with target events, K is smaller than or equal to N, screened data form an intermediate data list X', the intermediate data list is a list with M rows and K columns, each row represents an account to be sorted, and each column identifies one characteristic of the account.
S303, determining a characteristic value of a target characteristic in the intermediate data table, presetting a threshold interval, setting a corresponding weight value for each interval, determining a grouping interval where the target characteristic of each account is located in the preset threshold interval according to the characteristic value, adding the weight values of the target characteristic of each account to obtain a target value according to the grouping result, and taking the target value as a first probability of triggering a target event by the target account.
S304, obtaining the characteristic value of the target characteristic in the intermediate data table, selecting a trained machine learning model (such as trained LR, XGboost or DeepFM) and taking the characteristic value of the target characteristic of each account as input to output a model predicted value, and determining the model predicted value as a second probability of triggering a target event by the target account.
And S305, fusing the first probability and the second probability according to a predetermined fusion scheme to obtain a final third probability, wherein the fusion scheme can be calculating a weighted average of the first probability and the second probability, or calculating a geometric average of the first probability and the second probability.
S306, determining a target account with the maximum third probability in the plurality of accounts according to the third probability of each account.
Fig. 4 is a block diagram of an alternative scoring ranking according to an embodiment of the application, as shown in fig. 4:
raw data: the raw data is the first-hand data based on the perception device collecting or docking the customer traffic database.
Derivative data: the derived data is processed based on the original data, such as conversion into map data, time series data, spatio-temporal data, and the like.
Data normalization: based on the original data and the derived data, data sorting service standardization is performed, and the fields are required by the service, and a mapping relation, missing value processing of partial data and other cleaning processes are required to be made, such as: the field of the numeric type is standardized for a part of the data (e.g., a sex field, 1 for male and 0 for female) if a default value (e.g., 0 or 1) is not filled in or null, etc.
A rule engine: based on the business rules given by domain expert knowledge, the corresponding values can be analyzed and calculated. Such as: and the Harts rule engine aiming at the space-time big data processes the business rules given by the public security experts related to the technical and tactical laws.
An algorithm engine: the method mainly realizes algorithms of service scenes, such as various centrality metrics and node importance metrics of graph mining algorithms based on graph data, trajectory similarity calculation based on spatio-temporal data, and the like.
Multiple targets in a service scene: the multiple targets refer to decision influencing factors of the current business scene, and the calculation of the multiple target attribute values of the specific business scene is completed through a rule engine and an algorithm engine.
Learning of scoring function weight: the scoring function weights are learned based on a machine learning model and a user instance refinement method.
Grading and sorting: and calculating scores based on the scoring function, and then carrying out forward and reverse sorting according to the scores.
It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present application is not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required in this application.
Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but the former is a better implementation mode 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 is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling an electronic device (such as a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present application.
According to another aspect of the embodiments of the present application, there is also provided a target account determination apparatus for implementing the target account determination method. Fig. 5 is a schematic diagram of an alternative target account number determination apparatus according to an embodiment of the present application, and as shown in fig. 5, the apparatus may include:
a predicting module 52, configured to predict a first probability that each account in a plurality of accounts triggers a target event through an event prediction model, where the event prediction model is obtained by training an initial event prediction model using an account sample set labeled with a trigger tag, and the trigger tag is used to indicate whether an account sample in the account sample set triggers the target event;
a first determining module 54, configured to determine a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, where the second probability corresponding to each account is obtained by analyzing a usage history of each account corresponding to the target event;
the second determining module 56 determines, according to the third probability corresponding to each account, a target account having an association relationship with the target event from the plurality of accounts.
It should be noted that the prediction module 52 in this embodiment may be configured to execute the step S202 in this embodiment, the first determination module 54 in this embodiment may be configured to execute the step S204 in this embodiment, and the second determination module 56 in this embodiment may be configured to execute the step S206 in this embodiment.
It should be noted here that the modules described above are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the above embodiments. It should be noted that the modules described above as a part of the apparatus may operate in a hardware environment as shown in fig. 1, and may be implemented by software or hardware.
Through the module, the technical problem that the credibility of the account which is determined from the accounts and has the incidence relation with the event is low is solved, and the technical effect that the credibility of the account which is determined from the accounts and has the incidence relation with the event is low is achieved.
Optionally, the first determining module comprises one of: a first determining unit, configured to determine a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account; a second determining unit, configured to determine a geometric average of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account.
Optionally, the first determining unit is configured to: acquiring a first weight value corresponding to the event prediction model and a second weight value corresponding to an analysis process of account use history, wherein the first weight value is used for indicating the reliability of the event prediction model, and the second weight value is used for indicating the reliability of the analysis process of account use history; and summing and calculating the first weight value as the weight of the first probability of each account and the second weight value as the weight of the second probability of each account to obtain the third probability corresponding to each account.
Optionally, the apparatus further comprises: an obtaining module, configured to obtain account data generated in a using process of each account before determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account; an extraction module, configured to extract a feature value of a target attribute feature from the account data, where the target attribute feature is an attribute feature corresponding to the target event in all attribute features of the account data; and the analysis module is used for analyzing the characteristic value of the target attribute characteristic of each account to obtain the second probability corresponding to each account.
Optionally, the prediction module comprises: a first obtaining unit, configured to obtain a feature value of a target attribute feature included in each account, where the target attribute feature is an attribute feature corresponding to the target event in all attribute features of the account data; the input unit is used for inputting the characteristic value of the target attribute characteristic included in each account into the event prediction model; and the second acquisition unit is used for acquiring a target probability value output by the event prediction model as the first probability corresponding to each account.
Optionally, the apparatus further comprises: an input module, configured to input the account sample set into the initial event prediction model before inputting a feature value of a target attribute feature included in each account into the event prediction model; the adjusting module is used for adjusting the model parameters of the initial event prediction model according to the comparison value between the initial probability value output by the initial event prediction model and the trigger tag until the comparison value between the initial probability value and the trigger tag falls into a preset threshold range; and a third post-determination module, configured to determine, as the target model parameter of the event prediction model, a model parameter that makes a comparison value between the initial probability value and the trigger tag fall within a preset threshold range.
Optionally, the second determining module comprises one of: a third obtaining unit, configured to obtain, from the multiple accounts, a target number of accounts with a highest third probability as the target account; a fourth obtaining unit, configured to obtain, from the multiple accounts, an account with the third probability higher than a probability threshold as the target account.
It should be noted here that the modules described above are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the above embodiments. It should be noted that the modules described above as a part of the apparatus may be operated in a hardware environment as shown in fig. 1, and may be implemented by software, or may be implemented by hardware, where the hardware environment includes a network environment.
According to another aspect of the embodiment of the present application, there is also provided an electronic device for implementing the method for determining a target account.
Fig. 6 is a block diagram of an electronic device according to an embodiment of the present application, and as shown in fig. 6, the electronic device may include: one or more processors 601 (only one of which is shown), a memory 603, and a transmission 605. as shown in fig. 6, the electronic apparatus may further include an input-output device 607.
The memory 603 may be configured to store software programs and modules, such as program instructions/modules corresponding to the method and apparatus for determining a target account in the embodiment of the present application, and the processor 601 executes various functional applications and data processing by running the software programs and modules stored in the memory 603, that is, implementing the method for determining a target account. The memory 603 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 603 may further include memory located remotely from the processor 601, which may be connected to the electronic device through 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 above-mentioned transmission device 605 is used for receiving or sending data via a network, and may also be used for data transmission between a processor and a memory. Examples of the network may include a wired network and a wireless network. In one example, the transmission device 605 includes a Network adapter (NIC) that can be connected to a router via a Network cable and other Network devices to communicate with the internet or a local area Network. In one example, the transmission device 605 is a Radio Frequency (RF) module, which is used for communicating with the internet in a wireless manner.
Among them, the memory 603 is used to store an application program, in particular.
The processor 601 may call the application stored in the memory 603 through the transmission device 605 to perform the following steps: predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set marked with a trigger tag, and the trigger tag is used for indicating whether the account sample in the account sample set triggers the target event or not; determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history corresponding to the target event of each account; and determining a target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
The embodiment of the application provides a method and a device for determining a target account. In the process of determining a target account from a plurality of accounts, predicting a first probability corresponding to each account in the plurality of accounts through an event prediction model, wherein the model is obtained by training an initial event prediction model through an account sample set marked with a trigger label for indicating whether an account sample in the account sample set triggers a target event or not, so that the first probability can objectively and accurately represent the association between the account and the target event, determining a third probability corresponding to each account according to a second probability obtained by analyzing the use history of each account to represent the final probability of triggering the target event by each account, and since the third probability is obtained by combining the prediction and analysis processes of a plurality of dimensions, the third probability can more accurately represent the possibility of triggering the target event by the account, so that the target account which is determined according to the third probability and has an association relation with the target event is more credible, therefore, the technical effect of improving the reliability of the account which is determined from the plurality of accounts and has the incidence relation with the event is achieved, and the technical problem that the reliability of the account which is determined from the plurality of accounts and has the incidence relation with the event is low is solved.
Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment is not described herein again.
It will be understood by those skilled in the art that the structure shown in fig. 6 is merely an illustration, and the electronic device may be a smart phone (e.g., an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, and a Mobile Internet Device (MID), a PAD, etc. Fig. 6 is a diagram illustrating a structure of the electronic device. For example, the electronic device may also include more or fewer components (e.g., network interfaces, display devices, etc.) than shown in FIG. 6, or have a different configuration than shown in FIG. 6.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by a program for instructing hardware associated with an electronic device, where the program may be stored in a computer-readable storage medium, and the storage medium may include: flash disks, Read-Only memories (ROMs), Random Access Memories (RAMs), magnetic or optical disks, and the like.
Embodiments of the present application also provide a storage medium. Optionally, in this embodiment, the storage medium may be used to execute a program code of the method for determining the target account.
Optionally, in this embodiment, the storage medium may be located on at least one of a plurality of network devices in a network shown in the above embodiment.
Optionally, in this embodiment, the storage medium is configured to store program code for performing the following steps: predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set marked with a trigger tag, and the trigger tag is used for indicating whether the account sample in the account sample set triggers the target event or not; determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history corresponding to the target event of each account; and determining a target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment is not described herein again.
Optionally, in this embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
The integrated unit in the above embodiments, if implemented in the form of a software functional unit and sold or used as a separate product, may be stored in the above computer-readable storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or a part of or all or part of the technical solution contributing to the prior art may be embodied in the form of a software product stored in a storage medium, and including instructions for causing one or more computer devices (which may be personal computers, servers, network devices, or the like) to execute all or part of the steps of the method described in the embodiments of the present application.
In the above embodiments of the present application, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one type of division of logical functions, and there may be other divisions when actually implemented, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The foregoing is only a preferred embodiment of the present application and it should be noted that those skilled in the art can make several improvements and modifications without departing from the principle of the present application, and these improvements and modifications should also be considered as the protection scope of the present application.
Claims (10)
1. A method for determining a target account number, comprising:
predicting a first probability of triggering a target event by each account in a plurality of accounts through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model by using an account sample set marked with a trigger tag, and the trigger tag is used for indicating whether the target event is triggered by the account sample in the account sample set;
determining a third probability corresponding to each account according to the first probability corresponding to each account and the second probability corresponding to each account, wherein the second probability corresponding to each account is obtained by analyzing the use history of each account corresponding to the target event;
and determining a target account having an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
2. The method of claim 1, wherein determining the third probability for each account according to the first probability for each account and the second probability for each account respectively comprises one of:
determining a weighted sum of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account;
determining a geometric mean of the first probability corresponding to each account and the second probability corresponding to each account as the third probability corresponding to each account.
3. The method of claim 2, wherein determining the weighted sum of the first probability for each account and the second probability for each account as the third probability for each account comprises:
acquiring a first weight value corresponding to the event prediction model and a second weight value corresponding to an analysis process of account use history, wherein the first weight value is used for indicating the reliability of the event prediction model, and the second weight value is used for indicating the reliability of the analysis process of account use history;
and summing and calculating the first weight value as the weight of the first probability of each account and the second weight value as the weight of the second probability of each account to obtain the third probability corresponding to each account.
4. The method of claim 1, wherein before determining the third probability for each account according to the first probability for each account and the second probability for each account, the method further comprises:
acquiring account data generated in the using process of each account;
extracting a characteristic value of a target attribute characteristic from the account data, wherein the target attribute characteristic is an attribute characteristic corresponding to the target event in all attribute characteristics of the account data;
analyzing the characteristic value of the target attribute characteristic of each account to obtain the second probability corresponding to each account.
5. The method of claim 1, wherein predicting, by the event prediction model, the first probability that the each account of the plurality of accounts triggers the target event comprises:
acquiring a characteristic value of a target attribute characteristic included in each account, wherein the target attribute characteristic is an attribute characteristic corresponding to the target event in all attribute characteristics of the account data;
inputting a characteristic value of the target attribute feature included in each account into the event prediction model;
and acquiring a target probability value output by the event prediction model as the first probability corresponding to each account.
6. The method of claim 5, wherein before inputting the feature value of the target attribute feature included in each account into the event prediction model, the method further comprises:
inputting the account sample set into the initial event prediction model;
adjusting model parameters of the initial event prediction model according to a comparison value between an initial probability value output by the initial event prediction model and the trigger tag until the comparison value between the initial probability value and the trigger tag falls into a preset threshold range;
determining a model parameter which enables a comparison value between the initial probability value and the trigger tag to fall within a preset threshold range as a target model parameter of the event prediction model.
7. The method of claim 1, wherein determining a target account from the plurality of accounts that has an association relationship with the target event according to the third probability corresponding to each account comprises one of:
acquiring the account with the target number with the highest third probability from the plurality of accounts as the target account;
and acquiring the account with the third probability higher than the probability threshold from the plurality of accounts as the target account.
8. An apparatus for determining a target account, comprising:
the prediction module predicts a first probability that each account in a plurality of accounts triggers a target event through an event prediction model, wherein the event prediction model is obtained by training an initial event prediction model through an account sample set marked with a trigger label, and the trigger label is used for indicating whether the account sample in the account sample set triggers the target event;
a first determining module, configured to determine a third probability corresponding to each account according to the first probability corresponding to each account and a second probability corresponding to each account, where the second probability corresponding to each account is obtained by analyzing a usage history of each account corresponding to the target event;
and the second determining module is used for determining a target account which has an association relation with the target event from the plurality of accounts according to the third probability corresponding to each account.
9. A storage medium, characterized in that the storage medium comprises a stored program, wherein the program when executed performs the method of any of the preceding claims 1 to 7.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the method of any of the preceding claims 1 to 7 by means of the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110495273.8A CN113205188A (en) | 2021-05-07 | 2021-05-07 | Method and device for determining target account, storage medium and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110495273.8A CN113205188A (en) | 2021-05-07 | 2021-05-07 | Method and device for determining target account, storage medium and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113205188A true CN113205188A (en) | 2021-08-03 |
Family
ID=77030195
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110495273.8A Pending CN113205188A (en) | 2021-05-07 | 2021-05-07 | Method and device for determining target account, storage medium and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113205188A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6430305B1 (en) * | 1996-12-20 | 2002-08-06 | Synaptics, Incorporated | Identity verification methods |
| CN111538912A (en) * | 2020-07-07 | 2020-08-14 | 腾讯科技（深圳）有限公司 | Content recommendation method, device, equipment and readable storage medium |
| CN112422480A (en) * | 2019-08-22 | 2021-02-26 | 腾讯科技（深圳）有限公司 | Method and device for determining account attribute, storage medium and electronic device |
| CN112669113A (en) * | 2020-12-30 | 2021-04-16 | 微民保险代理有限公司 | Product recommendation method and device, storage medium and electronic device |
-
2021
- 2021-05-07 CN CN202110495273.8A patent/CN113205188A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6430305B1 (en) * | 1996-12-20 | 2002-08-06 | Synaptics, Incorporated | Identity verification methods |
| CN112422480A (en) * | 2019-08-22 | 2021-02-26 | 腾讯科技（深圳）有限公司 | Method and device for determining account attribute, storage medium and electronic device |
| CN111538912A (en) * | 2020-07-07 | 2020-08-14 | 腾讯科技（深圳）有限公司 | Content recommendation method, device, equipment and readable storage medium |
| CN112669113A (en) * | 2020-12-30 | 2021-04-16 | 微民保险代理有限公司 | Product recommendation method and device, storage medium and electronic device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110198310B (en) | Network behavior anti-cheating method and device and storage medium | |
| CN107798027B (en) | Information popularity prediction method, information recommendation method and device | |
| CN106992994A (en) | A kind of automatically-monitored method and system of cloud service | |
| US10218575B2 (en) | Provision, configuration and use of a telecommunications network | |
| CN111340380B (en) | Client resource allocation method, device and storage medium | |
| CN116305168B (en) | Multi-dimensional information security risk assessment method, system and storage medium | |
| CN113271322B (en) | Abnormal flow detection method and device, electronic equipment and storage medium | |
| CN112966014A (en) | Method and device for searching target object | |
| CN114399321A (en) | Business system stability analysis method, device and equipment | |
| CN116915710A (en) | Traffic early warning method, device, equipment and readable storage medium | |
| CN110457387B (en) | Method and related device applied to user tag determination in network | |
| CN106294406A (en) | A kind of method and apparatus accessing data for processing application | |
| CN114245185A (en) | Video recommendation method, model training method, device, electronic equipment and medium | |
| CN112269937A (en) | Method, system and device for calculating user similarity | |
| KR101462858B1 (en) | Methods for competency assessment of corporation for global business | |
| CN113205188A (en) | Method and device for determining target account, storage medium and electronic device | |
| CN115186759A (en) | Model training method and user classification method | |
| CN113283484A (en) | Improved feature selection method, device and storage medium | |
| CN113076450A (en) | Method and device for determining target recommendation list | |
| CN112131607B (en) | Resource data processing method and device, computer equipment and storage medium | |
| CN117951393B (en) | Social media monitoring system and method | |
| CN108255583B (en) | Application program comparison method and device | |
| CN108319449B (en) | Application program architecture determining method and device | |
| CN114416963A (en) | Information recommendation model training method and device, storage medium and computer equipment | |
| CN116823480A (en) | Abnormality detection processing 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 |