CN112184321A

CN112184321A - Method, device and system for quickly screening client data

Info

Publication number: CN112184321A
Application number: CN202011084847.4A
Authority: CN
Inventors: 郝晨曦
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-10-12
Filing date: 2020-10-12
Publication date: 2021-01-05

Abstract

The invention discloses a method, a device and a system for quickly screening customer data, wherein the method comprises the steps of obtaining a customer label; the client label comprises a plurality of values of the dimension attributes, and the values of the dimension attributes are represented in a coding form; performing logical AND operation on the client label and the template label bit by bit to obtain a result label; the values of a plurality of dimension attributes of the template label are corresponding target values, and the values of other dimension attributes are all filled with 0; and judging whether the label of the result is the same as the label of the template, if so, taking the client corresponding to the label of the client as the target client. According to the invention, the client label and the template label are subjected to bitwise logic and operation, and the operation result is judged, so that the client data meeting all preset conditions can be quickly screened out.

Description

Method, device and system for quickly screening client data

Technical Field

The invention relates to the technical field of data processing, in particular to a method, a device and a system for quickly screening client data.

Background

For Customer Relationship Management (Customer Relationship Management) software, data about a Customer is likely to be on the order of tens of millions, or even hundreds of millions.

When data analysis is performed, client information meeting conditions needs to be found out quickly according to certain query conditions. At this time, the conventional method generally stores the query result in a database, and then performs the query by constructing a complicated query statement (such as an SQL statement), so that the query efficiency is often low. Even if the query efficiency is improved by creating the index, the creation of the index requires additional database operations, which affects the efficiency of data writing, and in addition, because the query method is diversified, the index capable of adapting to all scenes cannot be created in advance.

At present, some commercial database analysis platforms generally need very expensive hardware resources and software resources, and the cost is very high.

Disclosure of Invention

The purpose of the invention is as follows: in order to overcome the defects in the prior art, the invention provides a method, a device and a system for quickly screening client data, which have the advantages of less required computing resources and high operation efficiency.

The technical scheme is as follows: in order to achieve the above object, the method for rapidly screening client data of the present invention includes a screening process, where the screening process includes:

acquiring a client label; the client label comprises a plurality of values of the dimension attributes, and the values of the dimension attributes are represented in a coding form;

performing logical AND operation on the client label and the template label bit by bit to obtain a result label; the values of a plurality of dimensional attributes of the template label are corresponding target values, and the values of other dimensional attributes are all filled with 0; and

and judging whether the result label is the same as the template label or not, and if so, taking the client ID corresponding to the client label as a target client.

Further, all of the customer tags are contained in a customer database, and the method further comprises:

dividing all the customer tags in the customer database into a plurality of data groups; and

synchronously executing the screening process for each data group;

the obtaining of the customer tag in the screening process includes:

the customer tag is obtained from the data set.

Further, the synchronously executing the screening process for each of the data groups further comprises:

and merging all the target clients generated by all the screening processes.

Further, the method further comprises a database construction process, wherein the database construction process comprises the following steps:

acquiring customer data;

converting the attribute data of each dimension in the client data into codes according to a preset corresponding relation;

splicing the codes corresponding to the dimension attributes in sequence to obtain the client label of the client;

and adding the customer tag into a customer database or updating the customer tag corresponding to the existing customer in the customer database.

Further, the representation form of the value of each dimension attribute of the customer label is binary coding or octal coding or decimal coding or hexadecimal coding.

A customer data rapid screening apparatus, comprising:

an obtaining module for obtaining a customer label; the client label comprises a plurality of values of the dimension attributes, and the values of the dimension attributes are represented in a coding form;

the operation module is used for performing logic AND operation on the client label and the template label bit by bit to obtain a result label; the values of a plurality of dimensional attributes of the template label are corresponding target values, and the values of other dimensional attributes are all filled with 0; and

and the judging module is used for judging whether the result label is the same as the template label or not, and if so, taking the client ID corresponding to the client label as a target client.

The customer data rapid screening system comprises a memory and a controller, wherein an executable program is stored in the memory, and the controller can run the executable program to realize the customer data rapid screening method.

Has the advantages that: according to the client data rapid screening method, the client data rapid screening device and the client data rapid screening system, the client tags and the template tags are subjected to bitwise logic and operation, and the operation results are judged, so that the client data meeting all preset conditions can be rapidly screened.

Drawings

Fig. 1 is a schematic flow chart of a database creation process according to a first embodiment of the present invention;

FIG. 2 is a schematic flow chart of a screening process according to a first embodiment of the present invention;

fig. 3 is a schematic diagram illustrating a module configuration of a client data fast screening apparatus according to a second embodiment of the present invention;

fig. 4 is a schematic diagram illustrating a configuration of a customer data rapid screening system according to a third embodiment of the present invention.

Detailed Description

The present invention will be further described with reference to the accompanying drawings.

Example one

The customer data rapid screening method is used for screening customer data in a customer database, and is implemented by a data analysis platform capable of accessing the customer database, so that the screened data meet preset conditions. The data of the customer is stored in a customer database in the form of a customer tag, the customer tag is a string of codes, the code form can be a common code form such as a binary code, an octal code, a decimal code, a hexadecimal code and the like, can also be other existing code forms, can also be a brand new code form which does not exist in the future along with the development and progress of the technology, and the code forms are all regarded as falling into the protection scope of the invention.

The representation of the customer label in the following embodiments is illustrated in binary code form.

The client tag is composed of a plurality of fields arranged in sequence, the code of each field represents an attribute of one dimension, for the attribute of each dimension, the coded values corresponding to different values of the attribute are different, the value of each attribute has a corresponding code value corresponding to the value, and the number of characters contained in the field corresponding to the attribute of the dimension is determined according to the number of possible value taking situations of the attribute of the dimension, such as: there are three possible values of attribute "gender" corresponding to a field in the client tag, that is: unknown, male and female, therefore, two character length fields should be set for the attribute corresponding to the dimension to represent the three attribute values, and the three value cases corresponding to the unknown, male and female are represented by 00, 01 and 10 respectively.

In actual use, the data of the client generally includes attributes of many dimensions, for example, in an embodiment, the correspondence between the attributes, the values of the attributes, and the corresponding codes included in the client data is shown in the following table:

TABLE 1

As can be seen, each piece of customer data includes attributes of 20 dimensions (does not include a customer ID), and based on the correspondence table, the following database creation process may be performed, as shown in fig. 1, where the database creation process includes the following steps S101 to S104:

step S101, obtaining client data;

in this step, the customer data may be manually input by the customer, for example, information manually input by the customer when the customer registers a certain website or APP, or may be automatically generated by the database analysis platform according to the use data of the customer, generally, the customer data is a mixture of the former two, the fixed attributes such as gender, age, address and the like are manually input by the customer, and data such as preferred goods, preferred shopping channels, preferred shopping time and the like are obtained by the database analysis platform through statistical algorithm. Here, the customer data is also a hand data, and the data of each dimension attribute is also corresponding specific data, such as: the sex is still male and female, and the marital status is the specific data of married and unmarried.

Returning to fig. 1, in step S102, converting the attribute data of each dimension in the client data into a code according to a preset corresponding relationship;

in this step, the specific data of each dimension attribute can be converted into an abstract code according to the table, where the preset corresponding relationship is the attribute, the value of each attribute, and the corresponding code corresponding relationship table.

Returning to fig. 1, step S103, sequentially splicing the codes corresponding to the dimensional attributes to obtain the client tag of the client;

in this step, the data may be spliced in the sequence shown in the above table, after the data is spliced, the obtained client tag is an 83-bit binary number, the 83-bit binary number includes all attributes of the client, each attribute is divided according to a field, and a value of a specific attribute corresponding to the field can be obtained by reading a specific field of the binary number.

For example: the customer labels of the customers with ID C10001 in table 1 are shown in table 2 below (the tail of the first table is connected to the head of the second table, which actually represents a continuous table, and the following tables 3 to 6 are the same only because the typesetting needs to be broken, and the description thereof is not repeated):

TABLE 2

The customer label for the customer with ID C10002 in table 1 is described in table 3 below:

TABLE 3

Returning to fig. 1, step S104 adds the customer tag to the customer database or updates the customer tag corresponding to the customer already existing in the customer database.

In this step, if the client is a new client, the ID of the client and the client tag are added to the client database, and if the client is an existing client, the client tag of the client is updated.

Based on the customer database, the customer data rapid screening method of the present invention further includes a screening process, as shown in fig. 2, the screening process includes the following steps S201 to S203:

step S201, obtaining a client label;

in this step, the customer tag is obtained from a customer database.

Step S202, carrying out logical AND operation on the client label and the template label bit by bit to obtain a result label; the values of a plurality of dimensional attributes of the template label are corresponding target values, and the values of other dimensional attributes are all filled with 0;

in this embodiment, the clients meeting the following conditions need to be screened from the client database:

(1) age between 31 and 35;

(2) belonging to a common member;

(3) the amount of shopping in the past month is within 1 to 3 thousand;

(4) the preferred group of purchased items is "garment-man";

(5) the first channel preferred for purchase is "online-jingdong".

Corresponding to the above condition, the value of the corresponding field in the template tag may be set to the value corresponding to the condition, e.g. corresponding to the above condition (1), the value of the four-digit number in the second field in the customer tag should be 0110, corresponding to the above condition (2), the value of the four-digit number in the ninth field in the customer tag should be 0001, and so on, and the values of the other non-mentioned attributes are all filled with 0, the template tags as shown in table 4 may be constructed:

TABLE 4

The template label can be constructed by a worker and input into the database analysis platform, or constructed by the database analysis platform according to business needs. In the former case, the worker generally participates in indirect construction, i.e., the condition is input, and the database analysis platform generates the template tag according to the condition. In the latter case, the database analysis platform may determine the screening condition according to the product location of the commodity to be recommended and generate the template tag according to the screening condition, or in another embodiment, the database analysis platform may count the common characteristics of customers who purchase a certain commodity according to the customer tags of the customers, thus determining the condition of the target customer and generating the template tag according to the condition.

When the template tag and the client tag are subjected to logic and operation bit by bit, the occupied computing resources are very little, and the running efficiency of the computer is very high, so that the computer can quickly finish the bit-by-bit logic and operation even if the number of client data in the client database is very large (for example, 1 hundred million).

In this step, the result labels obtained for the customer C10001 and the customer C10002 are respectively shown in the following table 5 and the following table 6:

TABLE 5

TABLE 6

As can be seen from table 5, the client label satisfying the condition is logically and-operated with the template label bit by bit, and the obtained result label is equal to the template label; as can be seen from table 6, the result label obtained after performing logical and operation on the customer label and the template label bit by bit is different from the template label for the customer label that does not satisfy the condition. Therefore, the client label can be screened and then the next step is carried out.

Returning to fig. 2, in step S203, it is determined whether the result tag is the same as the template tag, and if so, the client ID corresponding to the client tag is taken as the target client.

In this step, when the client tag is the target tag, the processing of adding the ID of the target client to the target client list or marking the ID of the target client may be performed in the subsequent step.

The steps S201 to S203 are performed in a loop, so as to perform the above operations on all data in the client database, until each piece of user data is traversed, and all client data is filtered, so that all target clients can be obtained.

Within the computer language, the efficiency of performing the above process is very high. During query, the ID of the client and the information of the client tag are all read into the memory, because the information amount is small, each user tag has only 83 binary digits, which is equivalent to 83 bits, so that massive data can be stored in the limited memory, and then through logic and operation, massive data can be filtered and analyzed in a very short time. And through the method, the client data meeting the conditions can be screened out at one time, the condition-by-condition screening is not needed (namely, after the screening is finished according to one condition, the screening is carried out again according to the second condition in the rest conditions, and the like), and the screening efficiency can be greatly improved.

In the above steps S201 to S203, if the whole customer database is screened item by item, the efficiency is obviously low, and in order to solve the problem, the customer data fast screening method of the present invention further includes the following steps S301 to S303:

step S301, dividing all the customer labels in the customer database into a plurality of data groups;

step S302, synchronously executing the screening process aiming at each data group;

step S303, merging all the target clients generated by all the screening processes.

Thus, the step of acquiring the client tag in step S201 specifically includes: the customer tag is obtained from the data set.

Because the operation resources needed for executing one screening process are less, the customer database is grouped and all data groups are concurrently screened, so that the operation capability of hardware can be fully utilized, and the operation efficiency is greatly improved. For example, 5 million customers to be processed can be divided into 50 parts, each part has 1 million data, and the data is processed respectively, and then the results are combined together after the processing is completed. Such concurrent operation will make the whole algorithm have stronger scalability in performance.

And (3) forming a first sub-library by the client labels of the target clients obtained by screening the whole client database according to the steps S201 to S203, subsequently further screening the first sub-library by using the updated template labels according to the steps S201 to S203 to obtain a second sub-library, and so on, further screening the previous round of results by updating the template labels, and refining the screening results through multiple rounds of screening so that the screening results are closer to the actual requirements.

Example two

The present invention further provides a client data fast filtering apparatus 400, as shown in fig. 4, which is a schematic diagram of program modules of the client data fast filtering apparatus 400, in this embodiment, the client data fast filtering apparatus 400 may include or be divided into one or more program modules, and the one or more program modules are stored in a storage medium and executed by one or more processors, so as to complete the present invention and implement the above-mentioned client data fast filtering method. The program module referred to in the embodiments of the present invention refers to a series of computer program instruction segments capable of performing specific functions, and is more suitable for describing the execution process of the client data rapid screening method in the storage medium than the program itself. The following description will specifically describe the functions of the program modules of the present embodiment:

an obtaining module 401, configured to obtain a client tag; the client label comprises a plurality of values of the dimension attributes, and the values of the dimension attributes are represented in a coding form;

an operation module 402, configured to perform a logical and operation on the customer tag and the template tag bit by bit to obtain a result tag; the values of a plurality of dimensional attributes of the template label are corresponding target values, and the values of other dimensional attributes are all filled with 0; and

a judging module 403, configured to judge whether the result tag is the same as the template tag, and if so, take the customer ID corresponding to the customer tag as the target customer.

The process of implementing the client data fast filtering method based on the client data fast filtering apparatus 400 is described in detail in the first embodiment, and is not described herein again.

EXAMPLE III

The invention also provides a customer data rapid screening system 700, as shown in fig. 4, which includes a memory 701 and a controller 702, wherein the memory 701 stores an executable program, and the controller 702 can run the executable program to implement the customer data rapid screening method. The method for implementing the client data fast screening by the controller 702 running the executable program is described in detail in the embodiment, and is not described herein again.

According to the client data rapid screening method, the client data rapid screening device and the client data rapid screening system, the client tags and the template tags are subjected to bitwise logic and operation, and the operation results are judged, so that the client data meeting all preset conditions can be rapidly screened.

The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.

Claims

1. A customer data rapid screening method is characterized by comprising a screening process, wherein the screening process comprises the following steps:

and judging whether the result label is the same as the template label or not, and if so, listing the client ID corresponding to the client label as the target client.

2. The method for rapid screening of customer data according to claim 1, wherein all of the customer tags are contained in a customer database, the method further comprising:

synchronously executing the screening process for each data group;

the obtaining of the customer tag in the screening process includes:

the customer tag is obtained from the data set.

3. The customer data rapid screening method according to claim 2, wherein the step of performing the screening process synchronously for each data group further comprises:

and merging all the target clients generated by all the screening processes.

4. The customer data rapid screening method according to claim 1, further comprising a database construction process, wherein the database construction process comprises:

acquiring customer data;

5. The customer data rapid screening method according to claim 1, wherein the value of each dimension attribute of the customer label is represented in binary code or octal code or decimal code or hexadecimal code.

6. A customer data rapid screening device is characterized by comprising:

7. A customer data rapid screening system comprising a memory and a controller, wherein the memory stores an executable program, and the controller is operable to execute the executable program to implement the customer data rapid screening method according to any one of claims 1 to 5.