CN110209687B

CN110209687B - Multi-dimensional attribution query method and device

Info

Publication number: CN110209687B
Application number: CN201810155674.7A
Authority: CN
Inventors: 洪超
Original assignee: Beijing Gridsum Technology Co Ltd
Current assignee: Beijing Gridsum Technology Co Ltd
Priority date: 2018-02-23
Filing date: 2018-02-23
Publication date: 2021-06-22
Anticipated expiration: 2038-02-23
Also published as: CN110209687A

Abstract

The application discloses a multi-dimensional attribution query method and device. The method comprises the following steps: a query template is created by adopting a structured query language, wherein the query template is used for calculating the proportion of converting the access behavior under the attribution dimension into a preset index, and the attribution dimension is an analyzing condition of the access behavior; and inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template. By the method and the device, the problem that the conversion condition of the access behaviors under different dimensions is difficult to query in real time according to the data source in the related technology is solved.

Description

Multi-dimensional attribution query method and device

Technical Field

The application relates to the field of internet data analysis, in particular to a multi-dimensional attribution query method and device.

Background

At present, in the release of promotion information in the internet era, for example, the release of online advertisements, promotion soft texts and the like is often performed through a plurality of media channels, and the release forms in each media channel are different, different types of users may have primary knowledge about products related to the promotion information through the promotion information of the media a, then click and perform deep knowledge when seeing the promotion information of the product of the media B, finally perform product search in the media C, enter the official website according to the search result, and generate a behavior of purchasing the product, there have been some methods in the prior art to evaluate the influence of each media channel in the promotion information conversion, but how to analyze the conversion condition in each dimension and perform real-time query, for example, query the conversion amount generated after users of different genders access to advertisements, the conversion amount generated after the users of different ages access the advertisement is inquired, and the like, and the adjustment of the release form of the promotion information, the production amount of the product and the improvement of the product have a plurality of influences.

Aiming at the problem that the conversion condition of the access behaviors under different dimensions is difficult to inquire according to a data source in real time in the related technology, an effective solution is not provided at present.

Disclosure of Invention

The application mainly aims to provide a multi-dimension attribution query method and device to solve the problem that in the related art, the conversion condition of access behaviors under different dimensions is difficult to query according to a data source in real time.

To achieve the above object, according to one aspect of the present application, there is provided a multi-dimensional attribution query method. The method comprises the following steps: adopting a structured query language to create a query template, wherein the query template is used for calculating the proportion of converting the access behavior into the preset index under the attribution dimension, and the conversion behavior is the behavior of converting the access behavior into the preset index; and inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template.

Further, creating a query template in a structured query language includes: an input module is created by adopting a structured query language, and basic data are obtained through the input module, wherein the basic data at least comprise an access sequence, the number of users who firstly have access behaviors under the access sequence, the number of users who have access behaviors under a target attribution dimension, and the number of users who have conversion behaviors under the target attribution dimension, and the conversion behaviors are behaviors which are obtained by converting the access behaviors into preset indexes; establishing a source matrix, a transformation matrix and a jump matrix according to basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in an access sequence, the transformation matrix is composed of the number of users who have conversion behaviors in target attribution dimensions in the access sequence, and the jump matrix is composed of jump probabilities of the access behaviors in the target attribution dimensions in the access sequence; and establishing a query template according to the source matrix, the conversion matrix and the skip matrix.

Further, the obtaining of the basic data through the input module includes: determining a target time period and a target attribution dimension; counting access session data in a target time period, wherein the access session data at least comprises: the method comprises the steps of judging whether the time of access behaviors, the number of users of the access behaviors, user attributes and the access behaviors are converted into preset indexes or not, wherein the user attributes at least comprise user IDs; generating an access sequence in the order of access time based on the user ID within a target time period; acquiring the number of users with access behaviors occurring for the first time under the access sequence based on the access sequence and the access session data; acquiring the number of users with access behaviors in the target attribution dimension and the number of users with conversion behaviors in the target attribution dimension based on the target attribution dimension and the access session data; and taking the access sequence, the number of users with access behaviors occurring for the first time under the access sequence, the number of users with access behaviors occurring under the target attribution dimension and the number of users with conversion behaviors occurring under the target attribution dimension as basic data.

Further, creating a query template from the source matrix, the transformation matrix, and the hop matrix comprises: splicing a source matrix character string, a skip matrix character string and a conversion matrix character string with preset characters respectively for the source matrix, the skip matrix and the conversion matrix; transmitting the source matrix character string, the skip matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain an attribution function, wherein the attribution function is used for calculating and returning the proportion of converting access behaviors under each attribution dimension into preset indexes; a query template is created according to the attribution function.

Further, after the query template queries the proportion of the access behavior converted into the preset index under each cause dimension, the method further comprises the following steps: obtaining a query result; and analyzing the query result.

Further, the building of the jump matrix according to the basic data comprises: calculating the probability of the user jumping from any one-time access behavior to the next-time access behavior after the one-time access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimensionality in the access sequence; and forming a jump matrix according to the probability that the user jumps from the one-time access behavior to the next-time access behavior after the one-time access behavior under the access sequence.

To achieve the above object, according to another aspect of the present application, there is provided a query device for multidimensional attribution. The device includes: the system comprises a creating unit, a searching unit and a processing unit, wherein the creating unit is used for creating a query template by adopting a structured query language, the query template is used for calculating the proportion of converting an access behavior into a preset index under attribution dimensionality, and the conversion behavior is a behavior of converting the access behavior into the preset index; and the query unit is used for querying the proportion of the access behavior converted into the preset index under the attribution dimension in the query template.

Further, the creating unit includes: the acquisition module is used for establishing an input module by adopting a structured query language and acquiring basic data through the input module, wherein the basic data at least comprises an access sequence, the number of users who firstly have access behaviors in the access sequence, the number of users who have access behaviors in a target attribution dimension and the number of users who have conversion behaviors in the target attribution dimension, and the conversion behaviors are behaviors which are obtained by converting the access behaviors into preset indexes; the matrix creating module is used for creating a source matrix, a conversion matrix and a jump matrix according to basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in an access sequence, the conversion matrix is composed of the number of users who have conversion behaviors in target attribution dimensions in the access sequence, and the jump matrix is composed of the jump probability of the access behaviors in the target attribution dimensions in the access sequence; and the creating module is used for creating a query template according to the source matrix, the conversion matrix and the skip matrix.

In order to achieve the above object, according to another aspect of the present application, there is provided a storage medium including a stored program, wherein the program performs any one of the above-described multi-dimensional attribution query methods.

To achieve the above object, according to another aspect of the present application, there is provided a processor for executing a program, wherein the program executes any one of the above query methods for multidimensional attribution.

Through the application, the following steps are adopted: a query template is created by adopting a structured query language, wherein the query template is used for calculating the proportion of converting the access behavior under the attribution dimension into a preset index, and the attribution dimension is an analyzing condition of the access behavior; the proportion of converting the access behaviors under the attribution dimensionality into the preset indexes is inquired in the inquiry template, and the problem that the conversion condition of the access behaviors under different dimensionalities is difficult to inquire according to the data source in real time in the related technology is solved. And then the effect of inquiring the conversion condition of the access behavior under different dimensions in real time according to the data source is achieved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application. In the drawings:

FIG. 1 is a flow diagram of a multi-dimensional attributed query method provided in accordance with an embodiment of the present application; and

fig. 2 is a schematic diagram of a multi-dimensional attribution querying device according to an embodiment of the present application.

Detailed Description

It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.

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 should be understood that the data so used may be interchanged under appropriate circumstances such that embodiments of the application described herein may be used. 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.

For convenience of description, some terms or expressions referred to in the embodiments of the present application are explained below:

SQL: the Query Language refers to a Structured Query Language (Structured Query Language), which is a database Query and programming Language and is used for accessing data and querying, updating and managing a relational database system; the structured query language allows a user to work on high-level data structures, does not require the user to specify a method for storing data, does not require the user to know a specific data storage manner, and can use the same structured query language as an interface for data input and management.

UDF: the method is characterized in that the method is a user-defined function which is an ordered set of T _ SQL statements, the statements are optimized and compiled in advance and can be called as a unit, and parameters can be introduced when the UDF is used to support various different return values.

The present invention is described below with reference to preferred implementation steps, and fig. 1 is a flowchart of a query method for multidimensional attribution according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:

step S101, a query template is created by adopting a structured query language, wherein the query template is used for calculating the proportion of converting access behaviors into preset indexes under attribution dimensions, and the attribution dimensions are analysis conditions of the access behaviors;

taking mobile phone advertisement delivery analysis as an example, in application, an input module can be built by utilizing SQL, relevant data of behaviors of different users accessing mobile phone advertisements of brand A through various channels is input, and the input module is used for acquiring basic data required by calculation; the method comprises the steps that a calculation module used for calculating the conversion of access behaviors under attribution dimensions into preset indexes is built by utilizing SQL, an SQL query template is built on the basis of an input module and the calculation module, after the SQL query template is built, only attribution dimension data needs to be changed in the input template, and the proportion of the conversion of the access behaviors under the attribution dimensions into the preset indexes can be calculated, for example, the attribution dimensions are males in user genders, after a male user is queried to access mobile phone advertisements of a brand, the proportion of the quantity of mobile phones of the brand purchased by orders of the male user in the access of mobile phones of the brand to the quantity of mobile phone advertisements of the brand of the user in the access of the brand A.

And S102, inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template.

The SQL query template can perform corresponding calculation only by inputting the most original data, the original data including different attribution dimensions are input into an input module of the SQL query template, the calculation module calculates the proportion of converting the access behavior under each attribution dimension into the preset index, and the SQL query template can return the proportion of converting the access behavior under each attribution dimension into the preset index.

Optionally, in the query method for multidimensional attribution provided by the embodiment of the present application, creating a query template using a structured query language includes: an input module is created by adopting a structured query language, and basic data are obtained through the input module, wherein the basic data at least comprise an access sequence, the number of users who firstly have access behaviors in the access sequence, the number of users who have access behaviors in a target attribution dimension, and the number of users who have conversion behaviors in the target attribution dimension, wherein the conversion behaviors are behaviors which convert the access behaviors into preset indexes; establishing a source matrix, a transformation matrix and a jump matrix according to basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in an access sequence, the transformation matrix is composed of the number of users who have conversion behaviors in target attribution dimensions in the access sequence, and the jump matrix is composed of jump probabilities of the access behaviors in the target attribution dimensions in the access sequence; and establishing a query template according to the source matrix, the conversion matrix and the skip matrix.

For example, the target attribution dimension is male in the gender of the user, the access behavior is that different users access the mobile phone advertisement of the brand A through various channels, and the preset index is the behavior that the user places an order to purchase the mobile phone of the brand after accessing the mobile phone advertisement of the brand A; an input module is created by adopting a structured query language, an access sequence is obtained through the input module, each user can jump to other channels to access the brand mobile phone advertisement after accessing the brand mobile phone advertisement through one channel, after repeated jump access, the brand mobile phone is purchased by placing an order, and the behavior that the user jumps among different channels according to the time sequence in a target time period to browse the brand mobile phone advertisement forms the access sequence; for example, the target time period is 2017, month 1 and day 1 to 2017, month 1 and day 30, in the time period, a part of users access the brand a mobile phone advertisement in the browser a first, then access the brand a mobile phone advertisement in the browser b, jump to the browser c after the browser b accesses to access the brand a mobile phone advertisement, purchase the brand mobile phone after the browser c accesses to access, and browse the advertisements from the browser a to the browser b to the browser c, so that an access sequence is formed. And acquiring the number of users who firstly have access behaviors in the access sequence, wherein the first access in the access sequence refers to the access behavior corresponding to the first access channel in the access sequence, and for the path from the browser a to the browser b and then to the browser c for advertisement browsing, the number of the users who access the mobile phone advertisement of the brand A through the browser a is the number of the users who firstly have access behaviors in the access sequence. Acquiring the number of male users accessing the mobile phone advertisement of the brand A through each channel, and acquiring the number of male users who purchase the mobile phone of the brand after placing an order after accessing the mobile phone advertisement of the brand A through each channel;

and establishing a source matrix, a conversion matrix and a skip matrix according to basic data, wherein as shown in table 1, the source matrix comprises three access channels of a browser, a browser b and a browser c, and the number of mobile phone advertisements of the brand A accessed through the channels for the first time is used as a matrix element to form the source matrix under a plurality of access sequences of the users with different genders in the table, wherein the three browsers of a, b and c are respectively used as first access channels. For example, n1 indicates the number of users who access brand a cellular phone advertisement through the a browser channel for the first time in a plurality of access sequences using the a browser as the first access channel, and n2 indicates the number of users who access brand a cellular phone advertisement through the b browser channel for the first time in a plurality of access sequences using the b browser as the first access channel.

TABLE 1

Sex	a browser	b browser	c browser
				Male sex	n1	n2	n3
Female with a view to preventing the formation of wrinkles	n4	n5	n6

As shown in table 2, the users of different genders in the table construct a conversion matrix by ordering and purchasing the behavior number of the brand mobile phone as a matrix element after accessing the brand mobile phone advertisement through each channel after accessing the brand mobile phone advertisement through a plurality of access sequences each having a, b, and c as a final access channel, for example, p1 indicates the number of male users who purchased the brand mobile phone after accessing the brand mobile phone advertisement from a browser, p4 indicates the number of female users who purchased the brand mobile phone after accessing the brand mobile phone advertisement from a browser, p2 indicates the number of male users who purchased the brand mobile phone after accessing the brand mobile phone advertisement from b browser, p5 indicates the number of male users who purchased the brand mobile phone after accessing the brand mobile phone advertisement from b browser, in an access sequence having b browser as a final access channel, the number of female users who purchased the mobile phone after accessing the brand a mobile phone advertisement from the b browser, p3 the number of male users who purchased the mobile phone after accessing the brand a mobile phone advertisement from the c browser in the access sequence of the access channel ending with the c browser, p6 the number of female users who purchased the mobile phone after accessing the brand a mobile phone advertisement from the c browser in the access sequence of the access channel ending with the c browser;

TABLE 2

Sex	a browser	b browser	c browser
				Male sex	p1	p2	p3
Female with a view to preventing the formation of wrinkles	p4	p5	p6

In a preset time period, if a user accesses the mobile phone advertisement of brand A in browser a first, then accesses the mobile phone advertisement of brand A in browser b, jumps to browser c after the browser b accesses the mobile phone advertisement of brand A, and purchases the mobile phone advertisement of brand A after the browser c accesses the mobile phone advertisement of brand AThe behavior that the brand mobile phone browses advertisements from a browser a to a browser b to a browser c and finally purchases the advertisements forms an access sequence, and in the access sequence, the number of male users accessing the brand A mobile phone advertisements through the browser a for the first time is m₀After leaving the a browser, m₀M among male users₁A male user then accesses brand a's cell phone advertisement through the b browser,

that is, the probability that the male user jumps from the browser a to the browser b when viewing the brand A mobile phone advertisement, after leaving the browser b, m1 male users among the male users are₂A male user then accesses brand a's cell phone advertisement through the c-browser,

that is, the probability that the male user jumps from the browser b to the browser c when viewing the mobile advertisement of brand A, the access sequence is ended, and the jump probability of the male user access under the access sequence is sequentially obtained

And

and

elements in the jump matrix in the access sequence are formed, and similarly, the jump matrix accessed by the female user under the access sequence can also be calculated; and calculating the jump probability of each access sequence under the target attribution dimension to obtain a jump matrix of each access sequence under the target attribution dimension.

And establishing a query template according to the source matrix, the conversion matrix and the skip matrix, wherein the query template is used for querying the proportion of users of different genders who make orders and buy the brand mobile phones after accessing the mobile phone advertisements of the brand A through various channels.

Optionally, in the query method for multidimensional attribution provided in the embodiment of the present application, the obtaining of the basic data through the input module includes: determining a target time period and a target attribution dimension; counting access session data in a target time period, wherein the access session data at least comprises: the method comprises the steps of judging whether the time of access behaviors, the number of users of the access behaviors, user attributes and the access behaviors are converted into preset indexes or not, wherein the user attributes at least comprise user IDs; generating an access sequence in the order of access time based on the user ID within a target time period; acquiring the number of users with access behaviors occurring for the first time under the access sequence based on the access sequence and the access session data; acquiring the number of users with access behaviors in the target attribution dimension and the number of users with conversion behaviors in the target attribution dimension based on the target attribution dimension and the access session data; and taking the access sequence, the number of users with access behaviors occurring for the first time under the access sequence, the number of users with access behaviors occurring under the target attribution dimension and the number of users with conversion behaviors occurring under the target attribution dimension as basic data.

For example, in the target time period, the behavior that the user skips and browses brand A mobile phone advertisements among different channels according to the time sequence forms an access sequence; for example, the target time period is between 1 month 1 day 2017 and 1 month 30 year 2017, the target attribution dimension is male in gender; during the period, access session data generated by accessing the brand a mobile phone advertisement through the browser by the user each time is counted, for example, 5 users access the brand a mobile phone advertisement through the brand a browser at 12 o' clock 1/1 in 2017, wherein 3 male users are 2 female users, and the access session data comprises the time when the access behavior occurs, the number of users who have the access behavior, the gender of the users and the ID of the users. In a target time period, 5 users firstly access the mobile phone advertisement of the brand A in the browser a, then 3 male users in the 5 users access the mobile phone advertisement of the brand A in the browser b, 2 male users jump to the browser c to access the mobile phone advertisement of the brand A after the browser b accesses, 1 male user purchases the mobile phone of the brand after the browser c accesses, and the behaviors of browsing the advertisement from the browser a to the browser b and then to the browser c and finally purchasing form an access sequence, wherein the number of the users accessing the mobile phone advertisement of the brand A in the browser a is the number of the users firstly having access behaviors in the access sequence. Acquiring that in the access sequence, 3 male users access the mobile phone advertisement of brand A through a browser a, 3 male users access the mobile phone advertisement of brand A through a browser b, 2 male users access the mobile phone advertisement of brand A through a browser c, and 1 male user purchases the mobile phone advertisement of brand A; then the access sequence, the number of the users who have access behaviors for the first time under the access sequence, 5, the number of the male users who have access behaviors for each time, and the number of the male users who purchase the mobile phone through the access sequence, 1, are all used as basic data.

Optionally, in the query method for multi-dimensional attribution provided in the embodiment of the present application, creating a query template according to the source matrix, the transformation matrix, and the hop matrix includes: splicing a source matrix character string, a skip matrix character string and a conversion matrix character string with preset characters respectively for the source matrix, the skip matrix and the conversion matrix; and transmitting the source matrix character string, the jump matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain an attribution function, wherein the attribution function is used for calculating and returning the proportion of converting the access behavior under each attribution dimension into a preset index.

For example, a preset function is created under a structured language:

create function get_attribution(string,string,string,int,boolean)returns string location'/user/hadoop-wd/hongchao/impala-udf-1.0.2-jar-with-dependencies.jar'symbol＝'com.gridsum.attribution.matrixremoval.AttributionModelUdf'；

checking whether the creation of the preset function is successful:

select

get_attribution('c1::0.2,,c3::0.5,,c2::0.3','c1::c2::0.12,,c2::c3::0.23,,c1::c3::0.13,,c3::c2::0.32','c1::0.2,,c3::0.5,,c2::0.3',5,true)

wd3.get_attribution('c1::0.2,,c3::0.5,,c2::0.3',

'c1::c2::0.12,,c2::c3::0.23,,c1::c3::0.13,,c3::c2::0.32','c1::0.2,,c3::0.5,,c2::0.3',5,true)|

c3::0.590194738640246,,c2::0.3069695934403826,,c1::0.10283566791937142,

Fetched 1row(s)in 0.03s

and respectively splicing the source matrix character string, the skip matrix character string and the conversion matrix character string by using preset characters.

startMatrixStr source matrix string;

-fromtoMatrixStr jumping matrix strings;

-conversionMatrixStr converting the matrix string;

respectively splicing the source matrix and the conversion matrix into a source matrix character string and a conversion matrix character string;

,scStr as(

select group_concat(concat(source,'::',cast(start_count as String)),',,')as startMatrixStr

,group_concat(concat(source,'::',cast(conversion_count as String)),',,')as conversionMatrixStr

from scMatrix)

-splicing the hop matrices into a hop matrix string;

,fromtoStr as(

select group_concat(concat(from_source,'::',to_source,'::',cast(jump_rate as String)),',,')as fromtoMatrixStr

from fromtoMatrix

)

transmitting the source matrix character string, the skip matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain a cause function:

select get_attribution(startMatrixStr,fromtoMatrixStr,conversionMatrixStr,3,true)

the attribution function under the structured query language has a return mechanism, and when attribution dimensions are input, the proportion of converting access behaviors into preset indexes under each attribution dimension is returned.

It should be noted that, to ensure that the attribution function returns data at millisecond level, 48 (current cluster server core number) concurrences of shared thread pools are added to the attribution function, and the maximum unique value of the attribution dimension is limited, for example, a user may repeatedly self-jump between channels to browse the mobile advertisement of brand a and not buy, the maximum jump time within a certain period is limited to 1000, if the self-jump time exceeds 1000 when the same user browses the mobile advertisement of brand a within the period, an error is automatically reported, and the data under the user ID does not participate in the attribution calculation any more.

A user may check the mobile phone advertisement of the brand A through only one channel, and then purchase the mobile phone advertisement without jumping to other channels to check the advertisement, at the moment, no jumping behavior exists, no jumping matrix exists, the proportion of purchasing behavior generated after the mobile phone advertisement of the brand A is accessed under each attribution dimension is directly calculated based on the source matrix and the conversion matrix, and the result is returned without calculation through an attribution function.

Optionally, in the multi-dimensional attribution query method provided in the embodiment of the present application, after querying, in the query template, a ratio of access behavior to a preset index under each attribution dimension, the method further includes: obtaining a query result; and analyzing the query result.

It should be noted that the attribution function returns a queried value, where the queried value is a ratio of access behaviors in each attribution dimension separated by a separator to be converted into a preset index. And analyzing the proportion of the access behavior under each attribution dimension converted into the preset index according to the return value, and further adjusting the advertisement putting strategy.

For example: the merchant puts mobile advertisements of brand A in a C1 browser, a C2 browser, a C3 browser, a C4 browser, a C5 browser and a C6 browser, the attribution dimension is an advertisement putting channel, real-time query can be realized, and if the attribution dimension input in the input module is a channel, the return value is as follows:

sql

wd3.get_attribution(startmatrixstr,fromtomatrixstr,conversionmatrixstr,3,true)

C0::0.7703299154155997,,C1::0.18709972960001542,,C2::0.03469119758625809,,

C3::0.006872883650811622,,C4::5.967497128190447E-4,,C5::3.6750779269917176E-4,,C

6::4.201624179692769E-5

the attribution function returns the proportion of purchasing the brand mobile phone by placing an order after the user who is separated by the separator accesses the mobile phone advertisement of the brand A through different browsers, and separates the attribution dimensions by ': ' separating the attribution dimensions from the proportion of converting the access behavior under each attribution dimension into a preset index, namely an attribution value, such as ' C0::0.7703299154155997 ' indicates that the proportion of the contribution of purchasing the brand mobile phone by placing an order after accessing the mobile phone advertisement of the brand A through a C0 browser is about 77.0%, ' C1::0.18709972960001542 ' indicates that the proportion of the contribution of purchasing the brand mobile phone by placing an order after accessing the mobile phone advertisement of the brand A through a C1 browser is about 18.7%, ' C2::0.03469119758625809 ' indicates that the proportion of the contribution of purchasing the order after accessing the mobile phone advertisement of the brand A through the user through a C2 browser and generating a purchasing behavior is about 3.4%, ' C3: ' 0.006872883650811622 ' indicates that the contribution of purchasing the brand mobile phone advertisement of the brand mobile phone after accessing the brand A through a C3 browser The ratio is about 0.69%, "C4: 5.967497128190447E-4 "shows that the contribution ratio of purchasing brand of cell phone is about 0.006% after accessing brand A cell phone advertisement through C4 browser," C5: 3.6750779269917176E-4 "shows that the contribution rate of purchasing brand of cell phone after accessing brand A cell phone advertisement through C5 browser is about 0.0037%," C6: 4.201624179692769E-5 "indicates that the contribution rate of purchasing brand of cell phone after accessing brand A cell phone advertisement through C6 browser is about 0.0004%, according to the obtained query result, the conversion rate of the mobile phone advertisement of the brand put in the C0 browser is higher, the conversion rate of putting the mobile phone advertisement of the brand in the C4, C5 and C6 browsers is low, therefore, the proportion of the mobile phone advertisements of the brand put in the C0 browser is increased, and the putting proportion in other browsers is adjusted.

If the attribution dimension input in the input module is the user gender, the attribution function returns the proportion that users with different genders separated by separators access the mobile phone advertisements of the brand A through various channels and place orders to purchase the mobile phones of the brand, the obtained query result is male:: 0.18 and female:: 0.43, which indicates that the proportion of the customers who place orders to purchase the mobile phones of the brand after the male users access the mobile phone advertisements of the brand A through various channels is 18%, and the proportion of the customers who purchase the mobile phones of the brand after the female users access the mobile phone advertisements of the brand A through various channels and access the mobile phone advertisements of the brand A through various channels is 43%; according to the obtained query result, the mobile phone of the brand is popular with female customers, and product positioning is adjusted.

Optionally, in the query method for multidimensional attribution provided in the embodiment of the present application, establishing a jump matrix according to the basic data includes: calculating the probability of the user jumping from any one-time access behavior to the next-time access behavior after the one-time access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimensionality in the access sequence; and forming a jump matrix according to the probability that the user jumps from any one-time access behavior to the next-time access behavior after the one-time access behavior in sequence under the access sequence.

For example, the probability that a user jumps from one access behavior to the next access behavior in sequence under an access sequence forms a jump matrix, that is, under a jump sequence implied by the access sequence, the probability that the user jumps from the first access behavior to the second access behavior is counted, the probability that the user jumps from the second access behavior to the third access behavior is counted, and so on until the probability that the user jumps from the previous access behavior of the last access behavior of the sequence to the last access behavior, and the probabilities form elements in the jump matrix under the access sequence.

For another example, when the skip matrix is constructed, in a preset time period, if the user accesses the brand a mobile phone advertisement in the browser a first, then accesses the brand a mobile phone advertisement in the browser b, and accesses the brand a mobile phone advertisement in the browser bThen, the user jumps to the browser c to access the mobile phone advertisement of the brand A, purchases the mobile phone of the brand after the browser c accesses, browses the advertisement from the browser a to the browser b to the browser c and finally purchases the advertisement to form an access sequence, the target attribution dimension is male in the gender of the user, in the access sequence, one-time access behavior is any one-time access behavior in the access sequence, for example, the behavior of accessing the mobile phone advertisement of the brand A through the browser a is one-time access behavior, and the number of the male in the behavior of accessing the mobile phone advertisement of the brand A through the browser a is counted as m₀After leaving the browser a, the m is counted₀M among male users₁A male user then accesses brand a's cell phone advertisement through the b browser,

that is, the probability that the male user jumps from the browser a to the browser b when viewing the brand A mobile phone advertisement is counted after leaving the browser b that m1 male users have m₂A male user then accesses brand a's cell phone advertisement through the c-browser,

i.e. the probability that a male user jumps from browser b to browser c when viewing brand a mobile advertisement,

and

and constructing elements in the jump matrix under the access sequence under the dimension of the target attribution.

According to the multi-dimensional attribution query method provided by the embodiment of the application, a query template is created by adopting a structured query language, wherein the query template is used for calculating the proportion of converting access behaviors into preset indexes under attribution dimensions, and the attribution dimensions are analysis conditions of the access behaviors; the proportion of converting the access behaviors under the attribution dimensionality into the preset indexes is inquired in the inquiry template, and the problem that the conversion condition of the access behaviors under different dimensionalities is difficult to inquire according to the data source in real time in the related technology is solved. By inputting the data source into the query template, the attribution dimension is determined, the probability of converting the access behavior into the preset index is queried in real time, and the effect of querying the conversion condition of the access behavior under different dimensions according to the data source in real time is further achieved.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer-executable instructions and that, although a logical order is illustrated in the flowcharts, in some cases, the steps illustrated or described may be performed in an order different than presented herein.

The embodiment of the present application further provides a query device for multidimensional attribution, and it should be noted that the query device for multidimensional attribution of the embodiment of the present application can be used for executing the query method for multidimensional attribution provided by the embodiment of the present application. The following describes a multidimensional attribution query device provided by the embodiment of the present application.

FIG. 2 is a schematic diagram of a multi-dimensional ascribed querying device according to an embodiment of the present application. As shown in fig. 2, the apparatus includes: creation unit 10 and query unit 20:

the system comprises a creating unit 10, a query module and a query module, wherein the creating unit is used for creating a query template by adopting a structured query language, the query template is used for calculating the proportion of converting access behaviors into preset indexes under attribution dimensions, and the conversion behaviors are the behaviors of converting the access behaviors into the preset indexes;

and the query unit 20 is configured to query, in the query template, a proportion of the access behavior converted into the preset index under the attribution dimension.

Optionally, in the query device for multidimensional attribution provided in the embodiment of the present application, the creating unit 10 includes: the acquisition module is used for establishing an input module by adopting a structured query language and acquiring basic data through the input module, wherein the basic data at least comprises an access sequence, the number of users who firstly have access behaviors under the access sequence, the number of users who have access behaviors under a target attribution dimension and the number of users who have conversion behaviors under the target attribution dimension, and the conversion behaviors are behaviors which are obtained by converting the access behaviors into preset indexes; the matrix creating module is used for creating a source matrix, a conversion matrix and a jump matrix according to basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in an access sequence, the conversion matrix is composed of the number of users who have conversion behaviors in target attribution dimensions in the access sequence, and the jump matrix is composed of the jump probability of the access behaviors in the target attribution dimensions in the access sequence; and establishing a query template according to the source matrix, the conversion matrix and the skip matrix.

Optionally, in the query device for multidimensional attribution provided in the embodiment of the present application, the obtaining module includes: a determination submodule for determining a target time period and a target attribution dimension; the statistic submodule is used for counting the access session data in the target time period, wherein the access session data at least comprises: the method comprises the steps of judging whether the time of access behaviors, the number of users of the access behaviors, user attributes and the access behaviors are converted into preset indexes or not, wherein the user attributes at least comprise user IDs; the generation submodule is used for generating an access sequence according to the access time sequence based on the user ID in the target time period; the first acquisition submodule is used for acquiring the number of users with access behaviors occurring for the first time under the access sequence based on the access sequence and the access session data; the second acquisition submodule is used for acquiring the number of users generating access behaviors in the target attribution dimension and the number of users generating conversion behaviors in the target attribution dimension based on the target attribution dimension and the access session data; and the setting submodule is used for taking the access sequence, the number of users with access behaviors occurring for the first time under the access sequence, the number of users with access behaviors occurring under the target attribution dimension and the number of users with conversion behaviors occurring under the target attribution dimension as basic data.

Optionally, in the query device for multidimensional attribution provided in the embodiment of the present application, the creating module includes: the splicing submodule is used for splicing the source matrix character string, the skip matrix character string and the conversion matrix character string by using preset characters respectively for the source matrix, the skip matrix and the conversion matrix; creating a submodule for transmitting the source matrix character string, the skip matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain an attribution function, wherein the attribution function is used for calculating and returning the proportion of converting the access behavior under each attribution dimension into a preset index; a query template is created according to the attribution function.

Optionally, in the query device for multidimensional attribution provided in the embodiment of the present application, the device further includes: the acquisition unit is used for acquiring a query result; and the analysis unit is used for analyzing the query result.

Optionally, in the query device for multidimensional attribution provided in the embodiment of the present application, the matrix creating module includes: the calculation submodule is used for calculating the probability of the user jumping from any one-time access behavior to the next-time access behavior after the one-time access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimensionality in the access sequence; and the jump matrix construction submodule is used for constructing a jump matrix according to the probability that the user jumps from one access behavior to the next access behavior after the access behavior in sequence under the access sequence.

According to the multi-dimension attribution query device provided by the embodiment of the application, the creating unit 10 is used for creating a query template by adopting a structured query language, wherein the query template is used for calculating the proportion of converting an access behavior under attribution dimensions into a preset index, and the conversion behavior is a behavior of converting the access behavior into the preset index; the query unit 20 queries the proportion of the access behavior in the attribution dimension to be converted into the preset index in the query template, and solves the problem that the conversion condition of the access behavior in different dimensions is difficult to query in real time according to the data source in the related art. By inputting the data source into the query template, the attribution dimension is determined, the probability of converting the access behavior into the preset index is queried in real time, and the effect of querying the conversion condition of the access behavior under different dimensions according to the data source in real time is further achieved.

The multi-dimension attribution inquiring device comprises a processor and a memory, wherein the creating unit 10, the inquiring unit 20 and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. The kernel can be set to be one or more, and the conversion condition of the access behavior under different dimensions can be inquired according to the data source in real time by adjusting the kernel parameters.

The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

An embodiment of the present invention provides a storage medium on which a program is stored, which, when executed by a processor, implements the multi-dimensional attribution query method.

The embodiment of the invention provides a processor, which is used for running a program, wherein the multidimensional attribution query method is executed when the program runs.

The embodiment of the invention provides equipment, which comprises a processor, a memory and a program which is stored on the memory and can run on the processor, wherein the processor executes the program and realizes the following steps: adopting a structured query language to create a query template, wherein the query template is used for calculating the proportion of converting the access behavior into the preset index under the attribution dimension, and the conversion behavior is the behavior of converting the access behavior into the preset index; and inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template.

Creating a query template in a structured query language includes: an input module is created by adopting a structured query language, and basic data are obtained through the input module, wherein the basic data at least comprise an access sequence, the number of users who firstly have access behaviors in the access sequence, the number of users who have access behaviors in a target attribution dimension, and the number of users who have conversion behaviors in the target attribution dimension, wherein the conversion behaviors are behaviors which convert the access behaviors into preset indexes; establishing a source matrix, a transformation matrix and a jump matrix according to basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in an access sequence, the transformation matrix is composed of the number of users who have conversion behaviors in target attribution dimensions in the access sequence, and the jump matrix is composed of jump probabilities of the access behaviors in the target attribution dimensions in the access sequence; and establishing a query template according to the source matrix, the conversion matrix and the skip matrix.

The obtaining of the basic data by the input module includes: determining a target time period and a target attribution dimension; counting access session data in a target time period, wherein the access session data at least comprises: the method comprises the steps of judging whether the time of access behaviors, the number of users of the access behaviors, user attributes and the access behaviors are converted into preset indexes or not, wherein the user attributes at least comprise user IDs; generating an access sequence in the order of access time based on the user ID within a target time period; acquiring the number of users with access behaviors occurring for the first time under the access sequence based on the access sequence and the access session data; acquiring the number of users with access behaviors in the target attribution dimension and the number of users with conversion behaviors in the target attribution dimension based on the target attribution dimension and the access session data; and taking the access sequence, the number of users with access behaviors occurring for the first time under the access sequence, the number of users with access behaviors occurring under the target attribution dimension and the number of users with conversion behaviors occurring under the target attribution dimension as basic data.

Creating a query template according to the source matrix, the transformation matrix and the skip matrix comprises: splicing a source matrix character string, a skip matrix character string and a conversion matrix character string with preset characters respectively for the source matrix, the skip matrix and the conversion matrix; transmitting the source matrix character string, the skip matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain an attribution function, wherein the attribution function is used for calculating and returning the proportion of converting access behaviors under each attribution dimension into preset indexes; a query template is created according to the attribution function.

After the proportion of the access behavior converted into the preset index under each cause dimension is queried in the query template, the method further comprises the following steps: obtaining a query result; and analyzing the query result.

Establishing the jump matrix according to the basic data comprises: calculating the probability of the user jumping from the current access behavior to the next access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimensionality in the access sequence; and forming a jump matrix according to the probability that the user jumps from one access behavior to the next access behavior in sequence under the access sequence. The device herein may be a server, a PC, a PAD, a mobile phone, etc.

The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device: adopting a structured query language to create a query template, wherein the query template is used for calculating the proportion of converting the access behavior into the preset index under the attribution dimension, and the conversion behavior is the behavior of converting the access behavior into the preset index; and inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template.

Establishing the jump matrix according to the basic data comprises: calculating the probability of the user jumping from the current access behavior to the next access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimensionality in the access sequence; and forming a jump matrix according to the probability that the user jumps from one access behavior to the next access behavior in sequence under the access sequence.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A multi-dimensional attribution query method is characterized by comprising the following steps:

adopting a structured query language to create a query template, wherein the query template is used for calculating the proportion of converting access behaviors into preset indexes under attribution dimensions, and the attribution dimensions are analysis conditions of the access behaviors;

inquiring the proportion of the access behavior converted into the preset index under the attribution dimension in the inquiry template;

wherein creating a query template using a structured query language comprises:

an input module is created by adopting a structured query language, and basic data is obtained through the input module, wherein the basic data at least comprises an access sequence, the number of users who firstly have access behaviors in the access sequence, the number of users who have access behaviors in a target attribution dimension, and the number of users who have conversion behaviors in the target attribution dimension, and the conversion behaviors are behaviors which are obtained by converting the access behaviors into preset indexes;

establishing a source matrix, a conversion matrix and a jump matrix according to the basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in the access sequence, the conversion matrix is composed of the number of users who have conversion behaviors in the target attribution dimension in the access sequence, and the jump matrix is composed of the jump probability of the access behaviors in the target attribution dimension in the access sequence;

and establishing the query template according to the source matrix, the conversion matrix and the skip matrix.

2. The method of claim 1, wherein obtaining the base data via the input module comprises:

determining a target time period and a target attribution dimension;

counting access session data in the target time period, wherein the access session data at least comprises: the method comprises the steps that time of access behaviors, the number of users of the access behaviors, user attributes and whether the access behaviors are converted into preset indexes or not, wherein the user attributes at least comprise user IDs;

generating the access sequence in the order of access time based on the user ID within the target time period;

acquiring the number of users with access behaviors occurring for the first time in the access sequence based on the access sequence and the access session data;

acquiring the number of users with access behaviors in the target attribution dimension and the number of users with conversion behaviors in the target attribution dimension based on the target attribution dimension and the access session data;

and taking the access sequence, the number of users with access behaviors occurring for the first time in the access sequence, the number of users with access behaviors occurring in the target attribution dimension and the number of users with conversion behaviors occurring in the target attribution dimension as the basic data.

3. The method of claim 1, wherein creating the query template from the source matrix, the translation matrix, and the hop matrix comprises:

splicing the source matrix, the skip matrix and the conversion matrix into a source matrix character string, a skip matrix character string and a conversion matrix character string respectively by adopting preset characters;

and transmitting the source matrix character string, the skip matrix character string and the conversion matrix character string into a preset function under a structured query language to obtain an attribution function, wherein the attribution function is used for calculating and returning the proportion of converting the access behavior under each attribution dimension into a preset index.

4. The method of claim 3, wherein after querying the query template for the proportion of access behavior converted into the preset index in each attribution dimension, the method further comprises:

obtaining a query result;

and analyzing the query result.

5. The method of claim 1, wherein building a hopping matrix from the base data comprises:

calculating the probability of the user jumping from any one-time access behavior to the next-time access behavior after the one-time access behavior according to the number of the users with the corresponding access behaviors under the target attribution dimension in the access sequence;

and forming the jump matrix according to the probability that the user jumps from the one-time access behavior to the next-time access behavior after the one-time access behavior in sequence under the access sequence.

6. A multidimensional attribution querying device, comprising:

the system comprises a creating unit, a searching unit and a processing unit, wherein the creating unit is used for creating a query template by adopting a structured query language, the query template is used for calculating the proportion of converting an access behavior into a preset index under an attribution dimension, and the attribution dimension is an analyzing condition of the access behavior;

the query unit is used for querying the proportion of the access behavior converted into the preset index under the attribution dimension in the query template;

wherein the creating unit includes: the system comprises an acquisition module, a display module and a conversion module, wherein the acquisition module is used for establishing an input module by adopting a structured query language and acquiring basic data through the input module, the basic data at least comprises an access sequence, the number of users who firstly have access behaviors in the access sequence, the number of users who have access behaviors in a target attribution dimension and the number of users who have conversion behaviors in the target attribution dimension, and the conversion behaviors are behaviors which are obtained by converting the access behaviors into preset indexes; the matrix creating module is used for creating a source matrix, a conversion matrix and a jump matrix according to the basic data, wherein the source matrix is composed of the number of users who have access behaviors for the first time in the access sequence, the conversion matrix is composed of the number of users who have conversion behaviors in the target attribution dimension in the access sequence, and the jump matrix is composed of the jump probability of the access behaviors in the target attribution dimension in the access sequence; and the creating module is used for creating the query template according to the source matrix, the conversion matrix and the skip matrix.

7. A storage medium characterized in that the storage medium includes a stored program, wherein the program executes the multidimensional attribution query method of any one of claims 1 to 5.

8. A processor, configured to run a program, wherein the program when running performs the multidimensional attribution query method of any one of claims 1 to 5.