Dynamic cycle life cycle management method and system based on crowd labels

Abstract

The invention discloses a dynamic cycle life cycle management method and system based on crowd labels. The invention relates to the technical field of user portraits. A label mapping relation is established through a background interface of a label product system, a unique label ID is distributed to each label, and corresponding label information is filled in. Enabling each label to correspond to a number bin upstream value table and a field, so as to automatically obtain label data subsequently; according to the method, full-life-cycle management of crowd labels from production to offline is realized in a dynamic column mapping label form, and the complexity and the cost of label management are reduced. Labels can be added or deleted without frequently modifying a wide table structure, and the flexibility and expandability of the system are improved. According to the wide table data model provided by the invention, multi-dimensional information related to the same entity is stored in a centralized manner, and the problems of data redundancy and repeated storage are avoided. The multiplexing mechanism of the label field further reduces the data storage demand, and reduces the storage cost.

Description

Technical Field

[0001] The present invention relates to the technical field of user portraits, and in particular to a crowd labeling system processing technology for user portraits, and in particular to a dynamic weekly life cycle management method and system based on crowd labels. Background Art

[0002] As a core data application product in the Internet industry, user portraits describe users in multiple dimensions and assign various labels to users to achieve accurate user group segmentation and personalized service push. Especially in the e-commerce field, user portraits usually focus on key information such as the user's age, work and living area, consumption ability, and products of interest.

[0003] When building a crowd labeling system based on user portraits, the design of the data architecture is crucial. At present, the more mature data architectures are mainly divided into two types: wide table model and high table model.

[0004] (1) Wide table model: Store one record for each user, and use multiple fields to store various tag data of the user. This model is usually deployed in analytical databases. It can easily select user groups by using the database's powerful association and filtering capabilities. For example, when defining the target population of a marketing campaign, you only need to use SQL query statements and combine the conditional judgment of multiple portrait tags to quickly filter out the user groups that meet the conditions.

[0005] The advantage of the wide table model is that it can fully utilize the computing power of the storage engine to achieve efficient crowd selection. However, its disadvantages are also obvious, that is, every time a tag is added or offline, the table structure needs to be modified, which not only increases the complexity of maintenance, but also after the tag is offline, the related tag data and table fields still need to be retained and calculated, further increasing the difficulty of later maintenance.

[0006] (2) High-table model: Each user and each tag stores a record, which includes the user, tag name, and tag value. This model is usually stored in a data engine that supports fast query of large amounts of data, such as Elasticsearch or HBase. When using the high-table model, the process of defining a multi-condition population package usually includes first selecting users who meet each tag, and then calculating the intersection or union through the data structure capabilities of the service program.

[0007] The advantage of the high-level table model is that its label data does not depend on the model of the application layer table. There is no need to pay attention to the changes in the table structure every time a label is added or offline, thereby improving the flexibility of the system. However, its disadvantages cannot be ignored, that is, the data storage cost is high, and in the big data scenario, the computing power is completely dependent on the performance of the service, and there is a computing bottleneck.

[0008] To this end, the present invention proposes a dynamic weekly life cycle management method and system based on crowd tags. Summary of the invention

[0009] In view of this, the present invention hopes to provide a method and system for dynamic lifecycle management based on crowd tags to solve or alleviate the technical problems existing in the prior art, namely, how to control the lifecycle of crowd tags from production to offline in the form of dynamic column mapping tags in the application scenario based on the wide table data model, reduce data storage costs, and make the computing power of big data scenarios not completely dependent on the performance of the service, thereby solving the computing bottleneck. The technical solution of the present invention is implemented as follows:

[0010] First, a dynamic weekly life cycle management method based on crowd tags:

[0011] 1. Overview:

[0012] The present invention is based on a wide table data model and implements the lifecycle management of crowd labels by means of dynamic column mapping. A user label wide table and a label mapping table are generated in the database, which are used to store the mapping relationship between user label data and labels respectively. A label mapping relationship is established through the background interface of the label product system, and a unique label ID is assigned to each label, which is associated with the value table and field upstream of the data warehouse. Next, an SQL query task is automatically generated according to the label mapping relationship, and label data is extracted from the data warehouse at a regular interval and stored in the user label wide table.

[0013] The selection conditions and label mapping relationships are passed in through the label service interface, and the corresponding SQL query statements are generated to filter out the user groups that meet the conditions from the user label wide table, and the query results are output as CSV format files to the business side for subsequent processing. In this way, the solution can efficiently manage the life cycle of crowd labels and provide flexible crowd selection and query functions.

[0014] (II) Technical solution:

[0015] In order to achieve the above technical objectives, the present invention selects to execute the following steps S1 to S4.

[0016] 2.1 Step S1, generate data table:

[0017] Generate a user tag wide table in the database; and generate a tag mapping table at the same time.

[0018] 2.1.1 Step S100, pre-processing:

[0019] Generate a wide table name, such as user_tags_wide_table;

[0020] Generate fields, including user ID and reserved tag fields (such as tag1, tag2, tag3...);

[0021] The data type of the reserved tag field is selected according to the nature of the tag value, such as VARCHAR or INT.

[0022] 2.1.2 Step S101, create a user tag wide table:

[0023] Use SQL statements to create a user tag wide table in the database; for example:

[0024]

[0025] 2.1.3 Step S102, generate a label mapping table structure:

[0026] Generate a mapping table name such as tag_mapping_table. Define fields, including:

[0027] (1) tag_id: unique identifier of a tag, data type such as INT or BIGINT;

[0028] (2) tag_name_cn: Chinese name of the tag, data type such as VARCHAR;

[0029] (3) tag_name_en: English name of the tag, data type such as VARCHAR;

[0030] (4) subscription_status: tag subscription status, data type such as BOOLEAN or INT;

[0031] (5) upstream_table_name: the name of the upstream value table of the data warehouse, the data type is VARCHAR;

[0032] (6) upstream_column_name: the name of the upstream value field in the data warehouse, the data type is VARCHAR;

[0033] 2.1.4 Step S103, create a label mapping table:

[0034] Use SQL statements to create a label mapping table in the database; for example:

[0035]

[0036] 2.2 Step S2, establish label mapping relationship:

[0037] Through the backend interface of the label product system, establish a label mapping relationship, assign a unique label ID to each label, and fill in the corresponding label information. Make each label correspond to an upstream value table and field in the data warehouse so that label data can be automatically obtained later.

[0038] 2.2.1 Step S200, create a new label mapping relationship:

[0039] Create a new tag mapping relationship; assign a unique tag ID to each tag to ensure the uniqueness of the tag. Fill in the Chinese and English names of the tag; set the subscription status of the tag, including whether it is enabled and whether it is public.

[0040] 2.2.2 Step S201: Configure the upstream value information of the data warehouse:

[0041] Specify the name of the upstream data warehouse value table corresponding to the label to ensure that the required label data can be extracted from it. Select or fill in the upstream data warehouse value field name, which should contain the data used to generate the label.

[0042] 2.2.3 Step S202, save the label mapping relationship:

[0043] Save the filled label information and data warehouse upstream value information to the database.

[0044] 2.3 Step S3, automatically generate tags to perform tasks:

[0045] According to the upstream value table and field of each tag in the data warehouse, the corresponding SQL query statement for extracting tag data from the data warehouse is generated. The generated SQL task is submitted to the task scheduling system, which is executed regularly and the calculation results are stored in the user tag wide table.

[0046] 2.3.1 Step S300, obtaining label mapping relationship:

[0047] Query all enabled and active label mapping relationships from the label mapping table. Extract the data warehouse upstream value table name and field name corresponding to each label.

[0048] 2.3.2 Step S301, generate SQL query statement:

[0049] For each tag, use aggregate functions or conditional statements to generate SQL query statements based on the upstream value table name and field name of the data warehouse.

[0050] 2.3.3 Step S302, constructing a task execution plan:

[0051] A task execution plan is created for each generated SQL query statement, including the frequency of task execution (such as daily, weekly, etc.) and the time window for task execution when the system load is below the preset threshold.

[0052] 2.3.4 Step S303, execute the task regularly and store the result:

[0053] Submit the constructed task execution plan to the task scheduling system and write the results into the user tag wide table.

[0054] The task scheduling system executes tasks according to the plan. During execution, the system runs SQL query statements to extract tag data from the data warehouse. The extracted data is written into the user tag wide table and associated with the user ID.

[0055] 2.4 Step S4, crowd selection and query:

[0056] The selection conditions and label mapping relationships are passed in through the label service interface; the label service generates the corresponding SQL query statement based on the passed in conditions and mapping relationships;

[0057] Execute SQL query statements to filter out qualified user groups from the user tag wide table;

[0058] The query results are exported as a CSV file and provided to the business party for subsequent processing.

[0059] 2.4.1 Step S400, receiving the selection conditions and label mapping relationship:

[0060] Receive the selection conditions and label mapping relationships passed in by the business party through the label service interface; construct SQL query statements based on the passed in selection conditions and label mapping relationships;

[0061] Use the tag ID in the tag mapping relationship and the corresponding user tag wide table field for matching.

[0062] Construct a WHERE clause to filter users based on selection criteria, such as tag values ​​or tag combinations.

[0063] 2.4.2 Step S401, execute SQL query statement:

[0064] Submit the generated SQL query statement to the database for execution, and then obtain the query results from the database; you can also perform necessary processing on the query results, such as deduplication and sorting.

[0065] 2.4.3 Step S402, output CSV format file:

[0066] Convert the processed query results into a CSV format file, including the user ID and tag value, and set the character encoding to ensure the correctness of the file content.

[0067] 2.4.4 Step S403, providing query result file:

[0068] Provide the CSV format file to the business party; record the log of file transfer for tracking and auditing.

[0069] (III) Mechanism for solving technical problems:

[0070] 3.1 The format of dynamic column mapping tag is:

[0071] Dynamic column mapping tags are one of the core mechanisms of the technical solution of the present invention. By reserving enough tag fields (such as tag1, tag2, tag3...) in the wide table and maintaining the correspondence between each tag and these reserved fields in the tag mapping table, dynamic management of tags is achieved. When a new tag needs to be added to the system, it is only necessary to add the corresponding entry in the tag mapping table without modifying the wide table structure. Similarly, when a tag is no longer needed, it only needs to be deleted or marked as offline in the tag mapping table, and the corresponding fields in the wide table can be reused by other new tags.

[0072] 3.2 Crowd tag lifecycle management:

[0073] The life cycle management of crowd tags includes the production, use, maintenance and offline stages of tags. In the technical solution of the present invention, the effective management and control of the life cycle of crowd tags can be achieved through the form of dynamic column mapping tags:

[0074] (1) Label production: The addition of new labels only requires configuration in the label mapping table, without modifying the wide table structure, thus reducing production costs.

[0075] (2) Tag usage: When selecting and querying a group of people, the corresponding SQL query statement is automatically generated based on the input selection criteria and tag mapping relationship, and user data that meets the criteria is extracted from the wide table, thereby improving query efficiency.

[0076] (3) Label maintenance: When the upstream value logic of a label changes, you only need to update the corresponding fields in the label mapping table to synchronize the label data in the wide table, reducing maintenance costs.

[0077] (4) Label offline: Labels that are no longer needed can be deleted or marked as offline in the label mapping table. The corresponding fields in the wide table can then be reused by other new labels, achieving effective resource recycling.

[0078] 3.3 Reduce data storage costs:

[0079] By dynamically mapping tags to columns, the technical solution of the present invention can significantly reduce data storage costs. Since the tag field in the wide table is reserved, the number of tags can be flexibly adjusted according to actual needs, avoiding the problem of table structure expansion caused by too many tags. At the same time, due to the reusability of tags, the data redundancy problem caused by tag offline is also reduced.

[0080] 3.4 Solving the computing bottleneck:

[0081] The technical solution of the present invention reduces the computing pressure of the service layer by centrally storing the crowd label data in a wide table and using the computing performance of the database itself to process query requests. In addition, by optimizing SQL query statements and database index strategies, query performance can be further improved to solve the computing bottleneck problem.

[0082] Second, a dynamic weekly life cycle management system based on crowd tags:

[0083] The system includes a processor and a memory connected to the processor, wherein program instructions are stored in the memory, and when the program instructions are executed by the processor, the processor executes the dynamic life cycle management method as described above, and the processor is connected to:

[0084] (1) A data table generation module for generating a user tag wide table and a tag mapping table in a database.

[0085] (2) A label mapping relationship management module for establishing, modifying and deleting label mapping relationships through the background interface of the label product system.

[0086] (3) A label data extraction and execution module that automatically generates SQL query statements based on label mapping relationships, extracts label data from the data warehouse, and executes these query tasks on a regular basis.

[0087] (4) A population selection and query module for receiving selection conditions and label mapping relationships through a label service interface, generating and executing SQL query statements, and filtering out user groups that meet the conditions from a user label wide table.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] 1. Efficient management of the life cycle of crowd tags: This invention realizes the full life cycle management of crowd tags from production to offline through the form of dynamic column mapping tags, reducing the complexity and cost of tag management. Tags can be added or deleted without frequent modification of wide table structures, which improves the flexibility and scalability of the system.

[0090] 2. Reduce data storage costs: The wide table data model provided by the present invention allows the multi-dimensional information related to the same entity to be stored centrally, avoiding the problem of data redundancy and repeated storage. The reuse mechanism of the tag field further reduces the data storage requirements and reduces the storage cost.

[0091] 3. Enhance the scalability and maintainability of the system: The modular design of the present invention makes it easier to add new functions and modify existing functions, reducing the difficulty of system maintenance. The dynamic column mapping mechanism makes the addition, deletion and modification of tags more flexible without the need for large-scale reconstruction of the system.

[0092] 4. Improve business response speed and flexibility: The label service interface of the present invention provides flexible circle selection and query functions, allowing the business party to quickly obtain the required population data. Its high efficiency enables the business party to respond to market changes and customer needs more quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0094] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0095] Figure 2 A schematic diagram of the mapping relationship between tags and data warehouses in the method of the present invention;

[0096] Figure 3 It is a schematic diagram of the system composition of the present invention. DETAILED DESCRIPTION

[0097] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below;

[0098] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0099] Explanation of relevant terms:

[0100] (1) Crowd tags: various descriptive information about the subject, such as the gender, age, spending power, income, and product types of consumers in e-commerce scenarios; and the order preferences, usual working hours, order-taking ability, and product stickiness of drivers in online car-hailing services.

[0101] (2) Lifecycle: defines the process from the birth to the death of a transaction. The lifecycle of a crowd tag specifically refers to the definition of the tag, the development of the tag, the launch of the tag, the use of the tag, the silencing of the tag, and the delisting of the tag.

[0102] (3) Population package: Use various label conditions to select the user groups that meet the conditions.

[0103] Embodiment 1: Figures 1-2 As shown, the embodiment will provide an application technology of a dynamic weekly life cycle management method based on crowd tags in the direction of online car-hailing customer operation and maintenance, including the following implementation plans.

[0104] In this embodiment, step S1: generating a data table includes:

[0105] Specifically, step S100: preprocessing: for the online car-hailing customer operation and maintenance scenario, generate a user tag wide table driver_tags_wide_table, the fields of which include driver_id (driver ID) and reserved tag fields tag1 to tagN, such as order acceptance preference, usual working hours, order acceptance ability, product stickiness, etc.

[0106] Specifically, step S101: create a user tag wide table: use an SQL statement to create driver_tags_wide_table.

[0107] Specifically, step S102: Generate a tag mapping table structure: Generate a mapping table

[0108] tag_mapping_table, the fields defined include tag_id, tag_name_cn (such as order preference), tag_name_en (such as Order Preference), subscription_status, upstream_table_name (such as driver_behavior_data), and upstream_column_name (such as preference_score).

[0109] Specifically, step S103: create a tag mapping table: use an SQL statement to create tag_mapping_table.

[0110] In this embodiment, step S2: establishing a label mapping relationship includes:

[0111] Specifically, step S200: create a new tag mapping relationship: create a new mapping relationship in the tag product system background, assign a unique ID to tags such as order preferences and customary working hours, fill in the Chinese name and English name, and set the subscription status.

[0112] Specifically, step S201: configure the data warehouse upstream value information: specify the order preference corresponding to the preference_score field of the driver_behavior_data table, the customary working hours corresponding to the working_hours field of the driver_schedule_data table, etc.

[0113] Specifically, step S202: save the label mapping relationship: save the label information and the data warehouse upstream value information to the database.

[0114] In this embodiment, step S3: automatically generating a tag execution task includes:

[0115] Specifically, step S300: obtain label mapping relationships: query all enabled and online label mapping relationships, and extract the data warehouse upstream value table name and field name.

[0116] Specifically, step S301: Generate SQL query statements: Generate SQL query statements for each tag, such as querying the preference_score field from the driver_behavior_data table to calculate the driver's order acceptance preference.

[0117] Specifically, step S302: constructing a task execution plan: creating a task execution plan for each SQL query statement, setting the execution frequency to daily, and the execution time to a period with low system load.

[0118] Specifically, step S303: Scheduled execution of tasks and storage of results: Submit the task execution plan to the task scheduling system, schedule the execution of SQL query statements, and write the results into

[0119] driver_tags_wide_table.

[0120] In this embodiment, step S4: crowd selection and query includes:

[0121] Specifically, step S400: receiving the selection conditions and the tag mapping relationship: receiving the selection conditions passed in by the business party through the tag service interface, such as drivers with high order acceptance preference, and constructing an SQL query statement.

[0122] Specifically, step S401: execute SQL query statement: execute SQL query statement to filter out the driver group that meets the conditions from driver_tags_wide_table.

[0123] Specifically, step S402: output CSV format file: convert the query results into a CSV format file, including information such as driver ID and order acceptance preference.

[0124] Specifically, step S403: providing a query result file: providing a CSV format file to the business party, and recording a file transfer log.

[0125] In this embodiment, the Python architecture program for the above solution is as follows:

[0126]

[0127]

[0128]

[0129]

[0130]

[0131] In the above program: SQLAlchemy library is used to define and create user tag wide table and tag mapping table. Pandas DataFrame is used to simulate inserting tag mapping relationship data into the tag mapping table. Data is extracted from the data warehouse, tag values ​​are calculated, and the results are stored in the user tag wide table. Finally, SQL query statements are constructed based on the given selection conditions, the query is executed, and the results are output as CSV format files. The query results can be provided to the business side through the API interface.

[0132] Embodiment 2: Figure 3As shown, based on the first embodiment, this embodiment further provides a dynamic weekly life cycle management system based on crowd tags:

[0133] The system includes a processor and a memory connected to the processor, wherein program instructions are stored in the memory, and when the program instructions are executed by the processor, the processor executes the dynamic life cycle management method as described above, and the processor is connected to:

[0134] (1) A data table generation module for generating a user tag wide table and a tag mapping table in a database:

[0135] (1.1) User tag wide table: contains user ID and reserved tag fields (such as tag1, tag2, tag3...), used to centrally store multi-dimensional tag information related to users.

[0136] (1.2) Label mapping table: contains fields such as label ID, label Chinese name, label English name, label subscription status, data warehouse upstream value table, and data warehouse upstream value field. It is used to maintain the correspondence between labels and reserved fields in the wide table, as well as the upstream data source information of labels.

[0137] (2) A label mapping relationship management module for establishing, modifying and deleting label mapping relationships through the backend interface of the label product system:

[0138] (2.1) Assign a unique tag ID to each tag and fill in the corresponding tag information (such as Chinese name, English name).

[0139] (2.2) Associate the tag with the upstream value table and field of the data warehouse so that the tag data can be automatically obtained later.

[0140] (2.3) Support offline operation of labels, that is, delete labels that are no longer needed in the label mapping table or mark them as offline.

[0141] (3) A label data extraction and execution module that automatically generates SQL query statements based on label mapping relationships, extracts label data from the data warehouse, and executes these query tasks on a regular basis:

[0142] (3.1) Generate the corresponding SQL query statement based on the upstream value table and field information of each tag.

[0143] (3.2) Submit the generated SQL task to the task scheduling system to ensure scheduled execution and store the calculation results in the user tag wide table.

[0144] (4) A crowd selection and query module for receiving the selection conditions and tag mapping relationships through the tag service interface, generating and executing SQL query statements, and filtering out user groups that meet the conditions from the user tag wide table:

[0145] (4.1) Supports flexible selection condition settings, such as single tag query, multiple tag combination query, etc.

[0146] (4.2) According to the input selection conditions and label mapping relationship, the corresponding SQL query statement is automatically generated.

[0147] (4.3) Execute SQL query statements and return query results, and support outputting the results to CSV format files or other formats for use by business parties.

[0148] All the above embodiments only express the implementation methods of the relevant practical applications of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

[0149] For those skilled in the art, it can be further appreciated that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0150] At the same time, those skilled in the art can understand that all or part of the processes in all the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

Claims

1. A dynamic weekly life cycle management method based on crowd tags, characterized in that: include: S1, generate a user tag wide table in the database; and generate a tag mapping table at the same time; S2, through the background interface of the label product system, establish the label mapping relationship, assign a unique label ID to each label, and fill in the corresponding label information; S3: Generate the corresponding SQL query statement for extracting label data from the data warehouse according to the upstream value table and field of each label; submit the generated SQL task to the task scheduling system, execute it regularly and store the calculation results in the user label wide table; S4, input the selection conditions and label mapping relationship through the label service interface; the label service generates the corresponding SQL query statement based on the input conditions and mapping relationship, and filters out the user groups that meet the conditions from the user label wide table; finally, the query results are provided to the business party.

2. The dynamic life cycle management method according to claim 1, characterized in that: The implementation method of S1 includes: S100, generating a wide table name and fields, including a user ID and a reserved tag field; S101, using SQL statements to create a user tag wide table in the database; S102, generating a mapping table name and defining fields; S103: Create a label mapping table in the database using an SQL statement.

3. The dynamic life cycle management method according to claim 2, characterized in that: In the S102, the fields include a unique tag identifier, a Chinese name of the tag, an English name of the tag, a tag subscription status, a data warehouse upstream value table name, and a data warehouse upstream value field name.

4. The dynamic life cycle management method according to claim 1, characterized in that: The implementation method of S2 includes: S200, creating a new tag mapping relationship; assigning a unique tag ID to each tag; S201, specify the data warehouse upstream value table name corresponding to the label, select or fill in the data warehouse upstream value field name, including the data used to generate the label; S202, save the filled label information and data warehouse upstream value information into the database.

5. The dynamic life cycle management method according to claim 1, characterized in that: The implementation method of S3 includes: S300: Query all enabled and active label mapping relationships from the label mapping table; extract the data warehouse upstream value table name and field name corresponding to each label; S301: For each tag, use an aggregate function or conditional statement to generate a SQL query statement based on the value table name and field name of its data warehouse upstream. S302, creating a task execution plan for each generated SQL query statement, including a frequency of task execution and a time window for task execution when the system load is below a preset threshold; S303, submitting the constructed task execution plan to the task scheduling system, and writing the result into the user tag wide table.

6. The dynamic life cycle management method according to claim 5, characterized in that: In the S3, the task scheduling system executes tasks according to the plan; during execution, SQL query statements are run to extract label data from the data warehouse; the extracted data is written into the user label wide table and associated with the user ID.

7. The dynamic life cycle management method according to claim 1, characterized in that: The implementation method of S4 includes: S400, receiving the selection conditions and label mapping relationship input by the business party through the label service interface; constructing an SQL query statement according to the input selection conditions and label mapping relationship; S401, submitting the generated SQL query statement to the database for execution, and then obtaining the query result from the database; S402, converting the processed query result into a file in CSV format, including the user ID and the tag value, and setting a character encoding for it; S403: Provide the file in CSV format to the business party and record the log of file transfer.

8. The dynamic life cycle management method according to claim 7, characterized in that: In the S400, the tag ID in the tag mapping relationship is matched with the corresponding user tag wide table field; a WHERE clause is constructed according to the tag value or tag combination to filter the user.

9. A dynamic weekly life cycle management system based on crowd tags, characterized by: The system includes a processor and a memory connected to the processor, wherein program instructions are stored in the memory. When the program instructions are executed by the processor, the processor executes the dynamic lifecycle management method as described in any one of claims 1-8.

10. The dynamic life cycle management system according to claim 9, characterized in that: The processor is connected with, A data table generation module for generating a user tag wide table and a tag mapping table in a database; A label mapping relationship management module for establishing, modifying and deleting label mapping relationships through the background interface of the label product system; The label data extraction and execution module is used to automatically generate SQL query statements based on label mapping relationships, extract label data from the data warehouse, and execute these query tasks regularly; A crowd selection and query module is used to receive the selection conditions and tag mapping relationships through the tag service interface, generate and execute SQL query statements, and filter out user groups that meet the conditions from the user tag wide table.

Images