CN114461630B - Smart attribution analysis method, device, equipment and storage medium - Google Patents

Abstract

The invention relates to an artificial intelligence technology, and discloses an intelligent attribution analysis method, which comprises the following steps: acquiring an initial data set and carrying out data replacement processing on the abnormal value to obtain a standard data set; dividing the standard data set into attribution phenomenon data sets and corresponding attribution factor data sets; importing the attribution phenomenon data set and the attribution factor data set into a preset model library, and calculating the prediction success rate of each model in the model library; determining an optimal attribution phenomenon prediction model according to the prediction success rate; and selecting a model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and utilizing the contribution degree of each attribution factor data of the model interpretation algorithm. In addition, the invention also relates to a blockchain technology, and the contribution degree of the initial data set and each attribution factor can be stored in nodes of the blockchain. The invention further provides an intelligent attribution analysis device, electronic equipment and a storage medium. The invention can improve the accuracy of attribution analysis.

Description

Intelligent attribution analysis method, device, equipment and storage medium

Technical Field

The present invention relates to the field of artificial intelligence technology, and in particular to an intelligent attribution analysis method, device, electronic device and computer-readable storage medium.

Background Art

Attribution analysis is an analytical method that explains which factors make up a phenomenon or effect. It is widely used in various industries such as the Internet advertising industry and the insurance industry. It is used to analyze the user behavior from which industry data comes and to improve the stickiness of industry users.

The current main attribution algorithms include rule-based attribution and data-driven attribution algorithms. Both require mathematical relationships to be set in advance. However, the business scenarios in actual industries are complex and changeable, and there is a lack of attribution analysis algorithms that are common to business scenarios in various industries, resulting in the low accuracy of existing attribution analysis methods.

Summary of the invention

The present invention provides an intelligent attribution analysis method, device and computer-readable storage medium, the main purpose of which is to solve the problem of low accuracy when performing attribution analysis.

To achieve the above object, the present invention provides an intelligent attribution analysis method, comprising:

Acquire an initial data set, perform data replacement processing on outliers in the initial data set, and obtain a standard data set;

Using a pre-built variable library, the standard data set is divided into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set;

Importing the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

Determining the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

A corresponding model interpretation algorithm is selected according to the type of the optimal attribution phenomenon prediction model, and the model interpretation algorithm is used to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

Optionally, performing data replacement processing on the outliers in the initial data set to obtain a standard data set includes:

Calculating a local reachable density ratio between each initial data in the initial data set and the neighborhood data of the initial data;

When the local density ratio is less than or equal to a preset density ratio threshold, determining that the initial data is an abnormal value;

The abnormal values are replaced with a preset correct data set to obtain a standard data set.

Optionally, the method of using a pre-built variable library to divide the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set includes:

Comparing the variable data in the standard data set with the pre-built variable library, determining the variable data in the standard data set that are consistent with those in the variable library as attributed phenomenon data, and determining the variable data in the standard data set that are inconsistent with those in the variable library as attribution factor data;

Calculating the correlation between the attributed phenomenon data and the attribution factor data, and determining the attribution factor data with the correlation greater than a preset threshold as the target attribution factor data corresponding to the attributed phenomenon data;

The attributed phenomenon data and the target attribution factor data are collected to obtain an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set.

Optionally, the calculating the correlation between the attributed phenomenon data and the attribution factor data includes:

Wherein, r(X,Y) is the correlation degree, X is the attributed phenomenon data, Y is the Yth attribution factor data, Cov(X,Y) is the covariance between the attributed phenomenon data and the attribution factor data, σ _x is the standard deviation of the attributed phenomenon data, and σ _y is the standard deviation of the attribution factor data.

Optionally, the calculating the prediction success rate of each attribution phenomenon prediction model in the to-be-attributed phenomenon prediction model library includes:

Dividing the attributed data set and the attribution factor data set into training samples and test samples according to a preset ratio;

Performing model training on each attribution phenomenon prediction model in the preset attribution phenomenon prediction model library according to the training samples to obtain multiple initial prediction models;

Using each of the initial prediction models to perform model prediction on the test samples to obtain test data for each of the initial test models;

Calculate the difference between the test data of each of the initial test models and the attributed phenomenon data of the test sample;

The test data whose difference is less than a preset threshold is determined as the correct prediction data, and the proportion of the correct prediction data in the test data of each of the initial test models is calculated to obtain the prediction success rate of each attribution phenomenon prediction model.

Optionally, the using the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data includes:

Calculating the standard deviation of each attribution factor data in the attribution factor data set, and determining the disturbance range of each attribution factor data according to the standard deviation;

Performing data perturbation on the data of each attribution factor according to the perturbation range to obtain new data of each attribution factor;

Based on the model interpretation algorithm, the new data of each attribution factor is used to train a target linear regression model, and the weights corresponding to the data of each attribution factor are obtained;

The respective attribution factor data are multiplied by the corresponding weight to obtain the contribution degree of the respective attribution factor data.

Optionally, the step of training a target linear regression model using the new data of each attribution factor based on the model interpretation algorithm includes:

Calculating the distance between each attribution factor data and the new data of each attribution factor respectively, and using the distance value as the weight of the new data of each attribution factor;

Using the optimal attribution phenomenon model to predict the attributed phenomenon for the new data of each attribution factor, to obtain the attributed phenomenon prediction data, and using the attributed phenomenon prediction data as the label data corresponding to the new data of each attribution factor;

Based on a preset model interpretation algorithm, the label data and new data of each attribution factor with weights are used to train a target linear regression model.

In order to solve the above problems, the present invention further provides an intelligent attribution analysis device, the device comprising:

A standard data set acquisition module is used to acquire an initial data set, perform data replacement processing on abnormal values in the initial data set, and obtain a standard data set;

A standard data set division module, used for dividing the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set by using a pre-built variable library;

A model prediction success rate calculation module, used to import the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculate the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

An optimal attribution phenomenon prediction model determination module, used to determine the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

The attribution factor data contribution calculation module is used to select the corresponding model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and use the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

at least one processor; and,

a memory communicatively connected to the at least one processor; wherein,

The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the intelligent attribution analysis method described above.

In order to solve the above problem, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned intelligent attribution analysis method.

The embodiment of the present invention divides the standard data set into the attributed phenomenon data set and the corresponding attribution factor data set through the pre-constructed variable library, and divides the data according to different industry variables, which is conducive to improving the accuracy of the data; the attributed phenomenon data set and the corresponding attribution factor data set are imported into the preset attribution phenomenon prediction model library, and the prediction success rate of each attribution phenomenon prediction model is calculated; then the optimal attribution phenomenon prediction model is determined from the attribution phenomenon prediction model library according to the prediction success rate, and the optimal attribution phenomenon prediction model that best matches the data is selected according to different data, so as to further improve the accuracy of the attribution analysis; the corresponding model interpretation algorithm is selected according to the type of the optimal attribution phenomenon prediction model, and the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data is calculated by using the model interpretation algorithm, so as to obtain a more accurate attribution analysis result. Therefore, the intelligent attribution analysis method, device, electronic device and computer-readable storage medium proposed by the present invention can solve the problem of low accuracy when performing attribution analysis.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a flow chart of an intelligent attribution analysis method provided by an embodiment of the present invention;

FIG2 is a schematic diagram of a process of dividing a standard data set provided by an embodiment of the present invention;

FIG3 is a schematic diagram of a flow chart of a calculation model predicting success rate according to an embodiment of the present invention;

FIG4 is a schematic diagram of a process for calculating the contribution of attribution factor data according to an embodiment of the present invention;

FIG5 is a functional module diagram of an intelligent attribution analysis device provided by an embodiment of the present invention;

FIG6 is a schematic diagram of the structure of an electronic device for implementing the intelligent attribution analysis method provided by an embodiment of the present invention.

The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings.

DETAILED DESCRIPTION

It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

The embodiment of the present application provides an intelligent attribution analysis method. The execution subject of the intelligent attribution analysis method includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the intelligent attribution analysis method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

1 is a flow chart of an intelligent attribution analysis method provided by an embodiment of the present invention. In this embodiment, the intelligent attribution analysis method includes:

S1. Obtain an initial data set, perform data replacement processing on outliers in the initial data set, and obtain a standard data set;

In an embodiment of the present invention, the initial data set is different business data sets of different industries, for example, relevant data of the Internet advertising industry, relevant data of the insurance industry, relevant data in an intelligent customer service scenario, etc.

In detail, the data replacement process is performed on the outliers in the initial data set to obtain a standard data set, including:

Calculating a local reachable density ratio between each initial data in the initial data set and the neighborhood data of the initial data;

When the local density ratio is less than or equal to a preset density ratio threshold, determining that the initial data is an abnormal value;

The abnormal values are replaced with a preset correct data set to obtain a standard data set.

In an embodiment of the present invention, the local reachable density is the inverse of the average distance from the neighborhood data of each initial data to the initial data. If the local reachable density ratio is greater than a preset density ratio threshold, it can be considered that the neighboring data and the initial data are in the same cluster and are not outliers.

Specifically, in the embodiment of the present invention, the local reachable density ratio of each initial data to the adjacent data in the initial data set is calculated using the following formula:

Wherein, LOF _k (P) is the local reachable density ratio, N _k (P) is the Pth initial data of the initial data set, ρ _k (P) is the local reachable density of the Pth data, ρ _k (O) is the average local reachable density of the neighborhood data O, and d _k (P, O) is the distance between the Pth initial data and the neighborhood data O.

In the embodiment of the present invention, data replacement processing is performed on the abnormal values in the initial data set to eliminate the abnormal data in the initial data set, thereby ensuring the rationality of the data in the initial data set and improving the accuracy of subsequent model selection.

S2, using a pre-built variable library to divide the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set;

In the embodiment of the present invention, the attributed phenomenon data set includes non-quantifiable elements, the attribution factor data set includes multiple possible factors affecting the attributed phenomenon data, and the pre-built variable library includes multiple attributed phenomenon data determined in advance.

In detail, referring to FIG. 2 , the method of using a pre-built variable library to divide the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set includes:

S21, comparing the variable data in the standard data set with the pre-built variable library, determining the variable data in the standard data set that are consistent with those in the variable library as attributed phenomenon data, and determining the variable data in the standard data set that are inconsistent with those in the variable library as attribution factor data;

S22, calculating the correlation between the attributed phenomenon data and the attribution factor data, and determining the attribution factor data with the correlation greater than a preset threshold as the target attribution factor data corresponding to the attributed phenomenon data;

S23. Collect the attributed phenomenon data and the target attribution factor data to obtain an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set.

Furthermore, in an embodiment of the present invention, the calculating of the correlation between the attributed phenomenon data and the attribution factor data includes:

Wherein, r(X,Y) is the correlation degree, X is the attributed phenomenon data, Y is the Yth attribution factor data, Cov(X,Y) is the covariance between the attributed phenomenon data and the attribution factor data, σ _x is the standard deviation of the attributed phenomenon data, and σ _y is the standard deviation of the attribution factor data.

Specifically, the attributed variables are different in different industry data, so the corresponding pre-built variable library should be selected according to the specific industry data of the embodiment, and the standard data set is divided into an attributed phenomenon data set and an attribution factor data set according to the pre-built variable library.

For example, in an embodiment of the present invention, in the advertising industry, the attributed variable may be advertising revenue, and the corresponding target attribution factor may be advertising click-through rate, user browsing time, etc.; in the insurance industry, the attributed variable may be renewal rate, and the corresponding target attribution factor may be the average insured age of customers, customer age composition, renewal rate of each insurance channel, etc.

S3, importing the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

In an embodiment of the present invention, the preset attribution phenomenon prediction model library may include hundreds of prediction models with prediction functions, including but not limited to time decay attribution model, Bayesian model, XGBoost, linear regression model, fully connected neural network and other models with prediction functions.

In detail, referring to FIG. 3 , the calculation of the prediction success rate of each attribution phenomenon prediction model in the to-be-attributed phenomenon prediction model library includes:

S31, dividing the attributed data set and the attribution factor data set into training samples and test samples according to a preset ratio;

S32, performing model training on each attribution phenomenon prediction model in the preset attribution phenomenon prediction model library according to the training samples to obtain multiple initial prediction models;

S33, using each of the initial prediction models to perform model prediction on the test sample to obtain test data for each of the initial test models;

S34, performing difference calculation between the test data of each of the initial test models and the attributed phenomenon data of the test sample;

S35. Determine the test data whose difference is less than a preset threshold as the correct prediction data, and calculate the proportion of the correct prediction data in the test data of each of the initial test models to obtain the prediction success rate of each attribution phenomenon prediction model.

Specifically, in an embodiment of the present invention, the preset threshold may be set to different thresholds according to different attributed phenomenon data sets. For example, if the attributed phenomenon data is a renewal rate, the threshold may be 0.01.

S4, determining the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

In the embodiment of the present invention, the attribution phenomenon prediction model that best matches the standard data set is selected from the preset attribution phenomenon prediction model library according to the model success rate, that is, the optimal attribution phenomenon prediction model.

For example, in an actual application scenario of the present invention, the preset attribution phenomenon prediction model library includes a time decay attribution model, a Bayesian model, XGBoost, a linear model, a fully connected neural network, etc., wherein the prediction success rate of the time decay attribution model is 89%, the prediction success rate of the Bayesian model is 85%, the prediction success rate of XGBoost is 89%, the prediction success rate of the linear model is 78%, and the prediction success rate of the fully connected neural network is 94%. The fully connected neural network is determined to be the optimal attribution phenomenon prediction model.

In an embodiment of the present invention, the optimal attribution phenomenon prediction model is determined by the model prediction success rate, and different attribution phenomenon prediction models can be selected according to different industry data, thereby improving the applicability of different industry data and ensuring the accuracy of subsequent attribution factor data contribution calculation.

S5. Select a corresponding model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and use the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

In an embodiment of the present invention, the corresponding model interpretation algorithm estimates the contribution of the variables, namely the attribution factors, in the optimal attribution phenomenon prediction model, that is, the role of each attribution factor data in the prediction result of the optimal attribution phenomenon prediction model.

Specifically, in the embodiments of the present invention, a model interpretation algorithm corresponding to the optimal attribution phenomenon prediction model can be called from a pre-stored model interpretation algorithm library according to the type of the optimal attribution phenomenon prediction model.

In detail, referring to FIG. 4 , the step of calculating the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data by using the model interpretation algorithm includes:

S51, calculating the standard deviation of each attribution factor data in the attribution factor data set, and determining the disturbance range of each attribution factor data according to the standard deviation;

S52, performing data perturbation on the data of each attribution factor according to the perturbation range to obtain new data of each attribution factor;

S53, based on the model interpretation algorithm, using the new data of each attribution factor to train a target linear regression model, and obtaining weights corresponding to the data of each attribution factor;

S54: multiply each attribution factor data by the corresponding weight to obtain the contribution of each attribution factor data.

Furthermore, in an embodiment of the present invention, the target linear regression model is obtained by training the new data of each attribution factor based on the model interpretation algorithm, including:

Calculating the distance between each attribution factor data and the new data of each attribution factor respectively, and using the distance value as the weight of the new data of each attribution factor;

Using the optimal attribution phenomenon model to predict the attributed phenomenon for the new data of each attribution factor, to obtain the attributed phenomenon prediction data, and using the attributed phenomenon prediction data as the label data corresponding to the new data of each attribution factor;

Based on a preset model interpretation algorithm, the label data and new data of each attribution factor with weights are used to train a target linear regression model.

For example, in an actual application scenario of the present invention, when the optimal attribution phenomenon prediction model is a fully connected neural network, the Deeplift algorithm can be called to interpret the optimal attribution phenomenon prediction model; when the optimal attribution phenomenon prediction model is an XGBoost model, the Shapley Value can be called to interpret the XGBoost model to obtain the contribution of each attribution factor in the optimal attribution phenomenon prediction model.

In an embodiment of the present invention, the contribution of each attribution factor data to the attributed phenomenon data is obtained through a model interpretation algorithm, and it is found out what kind of user behavior the industry data mainly comes from. According to the contribution of each attribution factor, a corresponding response strategy is made. For example, in the insurance industry data, it is found that the renewal rate is declining. The intelligent attribution analysis method is used to calculate that it is mainly caused by the increase in the complaint rate of old customers of a certain insurance channel.

The embodiment of the present invention divides the standard data set into the attributed phenomenon data set and the corresponding attribution factor data set through the pre-constructed variable library, and divides the data according to different industry variables, which is conducive to improving the accuracy of the data; the attributed phenomenon data set and the corresponding attribution factor data set are imported into the preset attribution phenomenon prediction model library, and the prediction success rate of each attribution phenomenon prediction model is calculated; then the optimal attribution phenomenon prediction model is determined from the attribution phenomenon prediction model library according to the prediction success rate, and the optimal attribution phenomenon prediction model that best matches the data is selected according to different data, so as to further improve the accuracy of the attribution analysis; the corresponding model interpretation algorithm is selected according to the type of the optimal attribution phenomenon prediction model, and the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data is calculated by using the model interpretation algorithm, so as to obtain a more accurate attribution analysis result. Therefore, the intelligent attribution analysis method proposed by the present invention can solve the problem of low accuracy when performing attribution analysis.

As shown in FIG5 , it is a functional module diagram of an intelligent attribution analysis device provided in one embodiment of the present invention.

The intelligent attribution analysis device 100 of the present invention can be installed in an electronic device. According to the functions to be implemented, the intelligent attribution analysis device 100 can include a standard data set acquisition module 101, a standard data set division module 102, a model prediction success rate calculation module 103, an optimal attribution phenomenon prediction model determination module 104 and an attribution factor data contribution calculation module 105. The module of the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

In this embodiment, the functions of each module/unit are as follows:

The standard data set acquisition module 101 is used to acquire an initial data set, perform data replacement processing on abnormal values in the initial data set, and obtain a standard data set;

The standard data set division module 102 is used to divide the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set by using a pre-built variable library;

The model prediction success rate calculation module 103 is used to import the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculate the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

The optimal attribution phenomenon prediction model determination module 104 is used to determine the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

The attribution factor data contribution calculation module 105 is used to select a corresponding model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and use the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

In detail, each module described in the intelligent attribution analysis device 100 described in the embodiment of the present invention adopts the same technical means as the intelligent attribution analysis method described in Figures 1 to 4 above when used, and can produce the same technical effects, which will not be repeated here.

As shown in FIG6 , it is a schematic diagram of the structure of an electronic device for implementing an intelligent attribution analysis method provided by an embodiment of the present invention.

The electronic device 1 may include a processor 10, a memory 11, a communication bus 12, and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as an intelligent attribution analysis program.

In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 11 (for example, executing intelligent attribution analysis programs, etc.), and calls data stored in the memory 11 to execute various functions of the electronic device and process data.

The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 11 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Furthermore, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 can be used not only to store application software and various types of data installed in the electronic device, such as the code of the intelligent attribution analysis program, but also to temporarily store data that has been output or is to be output.

The communication bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

The communication interface 13 is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

For example, although not shown, the electronic device may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

The intelligent attribution analysis program stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:

Acquire an initial data set, perform data replacement processing on outliers in the initial data set, and obtain a standard data set;

Using a pre-built variable library, the standard data set is divided into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set;

Importing the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

Determining the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

A corresponding model interpretation algorithm is selected according to the type of the optimal attribution phenomenon prediction model, and the model interpretation algorithm is used to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

Specifically, the specific implementation method of the processor 10 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

Furthermore, if the module/unit integrated in the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:

Acquire an initial data set, perform data replacement processing on outliers in the initial data set, and obtain a standard data set;

Using a pre-built variable library, the standard data set is divided into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set;

Importing the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library;

Determining the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate;

A corresponding model interpretation algorithm is selected according to the type of the optimal attribution phenomenon prediction model, and the model interpretation algorithm is used to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data.

In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims (7)

1. An intelligent attribution analysis method, characterized in that the method comprises: Acquire an initial data set, perform data replacement processing on outliers in the initial data set, and obtain a standard data set, wherein the initial data set includes data from the Internet advertising industry, data from the insurance industry, and/or data in an intelligent customer service scenario; Using a pre-built variable library, the standard data set is divided into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set; Importing the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library; Determining the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate; Selecting a corresponding model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and using the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data; The method of using a pre-built variable library to divide the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set includes: comparing the variable data in the standard data set with the pre-built variable library, determining the variable data in the standard data set that are consistent with those in the variable library as the attributed phenomenon data, and determining the variable data in the standard data set that are inconsistent with those in the variable library as the attribution factor data; calculating the correlation between the attributed phenomenon data and the attribution factor data, and determining the attribution factor data with the correlation greater than a preset threshold as the target attribution factor data corresponding to the attributed phenomenon data; and collecting the attributed phenomenon data and the target attribution factor data to obtain the attributed phenomenon data set and the attribution factor data set corresponding to the attributed phenomenon data set; The using of the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data includes: calculating the standard deviation of each attribution factor data in the attribution factor data set, and determining the disturbance range of each attribution factor data according to the standard deviation; performing data perturbation on each attribution factor data according to the disturbance range to obtain new data of each attribution factor; training a target linear regression model based on the model interpretation algorithm using the new data of each attribution factor to obtain the weight corresponding to each attribution factor data; multiplying each attribution factor data by the corresponding weight to obtain the contribution of each attribution factor data; The method of obtaining a target linear regression model by training the new data of each attribution factor based on the model interpretation algorithm includes: respectively calculating the distance between the data of each attribution factor and the new data of each attribution factor, and using the distance as the weight of the new data of each attribution factor; using the optimal attribution phenomenon prediction model to predict the attributed phenomenon for the new data of each attribution factor to obtain the attributed phenomenon prediction data, and using the attributed phenomenon prediction data as the label data corresponding to the new data of each attribution factor; and obtaining a target linear regression model by training the label data and the new data of each attribution factor with weights based on the preset model interpretation algorithm. 2. The intelligent attribution analysis method according to claim 1, characterized in that the step of performing data replacement processing on the outliers in the initial data set to obtain a standard data set comprises: Calculating a local reachable density ratio between each initial data in the initial data set and the neighborhood data of the initial data; When the local reachable density ratio is less than or equal to a preset density ratio threshold, determining that the initial data is an abnormal value; The abnormal values are replaced with a preset correct data set to obtain a standard data set. 3. The intelligent attribution analysis method according to claim 1, wherein the step of calculating the correlation between the attributed phenomenon data and the attribution factor data comprises: in, is the correlation degree, is the attributed phenomenon data, For the Attribution factor data, is the covariance between the attributed phenomenon data and the attribution factor data, is the standard deviation of the attributed phenomenon data, is the standard deviation of the attribution factor data. 4. The intelligent attribution analysis method according to claim 1, wherein the step of calculating the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library comprises: Dividing the attributed phenomenon data set and the attribution factor data set into training samples and test samples according to a preset ratio; Performing model training on each attribution phenomenon prediction model in the preset attribution phenomenon prediction model library according to the training samples to obtain multiple initial prediction models; Using each of the initial prediction models to perform model prediction on the test sample to obtain test data for each of the initial prediction models; Performing difference calculation between the test data of each of the initial prediction models and the attributed phenomenon data of the test sample; The test data whose difference is less than a preset threshold is determined as the correct prediction data, and the proportion of the correct prediction data in the test data of each of the initial prediction models is calculated to obtain the prediction success rate of each attribution phenomenon prediction model. 5. An intelligent attribution analysis device, used to implement the intelligent attribution analysis method according to any one of claims 1 to 4, characterized in that the device comprises: A standard data set acquisition module is used to acquire an initial data set, perform data replacement processing on abnormal values in the initial data set, and obtain a standard data set; A standard data set division module, used for dividing the standard data set into an attributed phenomenon data set and an attribution factor data set corresponding to the attributed phenomenon data set by using a pre-built variable library; A model prediction success rate calculation module, used to import the attributed phenomenon data set and the attribution factor data set into a preset attribution phenomenon prediction model library, and calculate the prediction success rate of each attribution phenomenon prediction model in the attribution phenomenon prediction model library; An optimal attribution phenomenon prediction model determination module, used to determine the optimal attribution phenomenon prediction model from the attribution phenomenon prediction model library according to the model prediction success rate; The attribution factor data contribution calculation module is used to select the corresponding model interpretation algorithm according to the type of the optimal attribution phenomenon prediction model, and use the model interpretation algorithm to calculate the contribution of each attribution factor data in the attribution factor data set to the attributed phenomenon data. 6. An electronic device, characterized in that the electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the intelligent attribution analysis method as described in any one of claims 1 to 4. 7. A computer-readable storage medium storing a computer program, characterized in that when the computer program is executed by a processor, the intelligent attribution analysis method as described in any one of claims 1 to 4 is implemented.