Adaptive table granularity data generation method and device, equipment and storage medium

Abstract

The present application relates to artificial intelligence and digital financial technology, and discloses a self-adaptive table granularity data generation method, device, equipment and storage medium in the data exploration scene of financial big data. The method comprises: performing index importance sorting on original data to obtain a business important dimension order table and obtain an initialization dimension; performing table granularity analysis on the original data according to the initialization dimension to obtain a divergence condition of the original data; using a pre-trained drilling analysis decision model to determine whether to perform drilling analysis on the original data according to the divergence condition; when the original data does not need to be drilled, the initialization dimension is configured as a self-adaptive table granularity, and data cleaning is performed on the original data according to the self-adaptive table granularity to generate self-adaptive table granularity data. The present application can adaptively identify the table granularity suitable for the business problem in the financial big data data exploration scene, and generate more scientific and reasonable data to be analyzed.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and digital finance, and in particular to an adaptive table-level data generation method, apparatus, device, and computer-readable storage medium for data exploration scenarios in financial big data. Background Technology

[0002] Currently, data exploration is required before any business analysis in the financial industry, such as loan analysis or customer intent analysis. However, when exploring source data, it is often found that after multiple iterations, the source system's tables contain data at multiple granularities, which poses a significant challenge for data warehouse personnel in determining the granularity and primary keys of the tables.

[0003] Currently, data exploration primarily involves analysts defining and analyzing dimensions, then reviewing the results and subjectively deciding whether to add more dimensions for drill-down analysis. The dimensions are then configured at the table level, and the data is cleaned to obtain the corresponding table-level data. This process is time-consuming and labor-intensive, the accuracy of the configured table granularity is often inaccurate, and the analysis is highly subjective, making it unsuitable for other analysts to use. Summary of the Invention

[0004] This invention provides a data generation method, apparatus, device, and storage medium with adaptive table granularity. Its main purpose is to generate more scientific and reasonable data to be analyzed by adaptively identifying the appropriate table granularity for business problems in financial big data exploration scenarios.

[0005] To achieve the above objectives, the present invention provides an adaptive table-granularity data generation method, comprising:

[0006] Step A: Obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business importance dimension order table, and obtain the initial dimension according to the business importance dimension order table;

[0007] Step B: Perform table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension;

[0008] Step C: Using a pre-trained drill-down analysis decision model, determine whether to perform drill-down analysis on the original data based on the divergence situation;

[0009] When the determination result indicates that drill-down analysis of the original data is required, step D: add dimensions to the initial dimensions according to the preset business importance dimension order table to obtain the updated dimensions, and return to the operation of step B, and perform table granular analysis on the original data according to the updated dimensions;

[0010] When the determination result is that drill-down analysis of the original data is not required, step E: configure the initial dimension as the adaptive table granularity, and perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0011] Optionally, the step of performing table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension includes:

[0012] Obtain the first part of the original data with a total number of statistical records greater than 1 under the initialization dimension, and obtain the number of records and creation time of the first part of the original data;

[0013] Obtain the second part of the original data where the number of summary statistics entries under the initialization dimension is equal to 1, and obtain the number of entries and creation time of the second part of the original data;

[0014] Based on the number of records and creation time of the first part of the original data, a histogram is constructed to obtain the data distribution; and based on the number of records and creation time of the second part of the original data, a line chart is constructed to obtain the data change trend.

[0015] The data distribution and the data change trend are combined and output to obtain the divergence of the original data under the initialization dimension.

[0016] Optionally, the step of using a pre-trained drill-down analysis decision model to determine whether to perform drill-down analysis on the original data based on the divergence situation includes:

[0017] The divergence features are extracted from the divergence situation using a pre-trained drill-down analysis decision model to obtain a divergence feature sequence;

[0018] Based on the initialization dimension, feature engineering is performed on the original data to extract features, resulting in a business feature sequence;

[0019] Based on the business feature sequence, determine the business problem complexity of the initialization dimension, and based on the divergent feature sequence, determine whether the dispersion condition satisfies the business problem complexity.

[0020] When the dispersion situation cannot meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

[0021] Optionally, the step of obtaining raw data from a pre-built business system cluster and sorting the raw data by indicator importance to obtain a business importance dimension order table includes:

[0022] Based on the preset business requirements, a cluster of business systems is selected from the pre-built enterprise main system;

[0023] The raw data is obtained from the business system cluster, and principal component analysis is performed on the raw data to obtain a set of principal component fields.

[0024] The principal component field set is fine-tuned according to the pre-built expert configuration labels to obtain a business importance dimension order table.

[0025] Optionally, before utilizing the pre-trained drill-down analysis decision model, the method further includes:

[0026] Obtain the pre-built initial binary classification decision model and the pre-processed business scenario data set;

[0027] Feature engineering is performed on the business scenario data set to extract training sample sets;

[0028] Target samples are extracted sequentially from the training sample set, and the initial binary classification decision model is used to perform drill-down analysis on the target samples to obtain prediction results.

[0029] Using the cross-entropy loss algorithm, the loss value between the true label corresponding to the target sample and the prediction result is calculated, and the loss value is minimized to obtain the model parameters when the loss value is minimized. Based on the model parameters, the initial binary classification decision model is updated by reverse network to obtain the updated binary classification decision model.

[0030] Record the loss values, obtain the loss value curve, and determine the convergence of the loss value curve;

[0031] When the loss curve fails to converge, return to the above operation of extracting target samples from the training sample set in sequence, and iteratively optimize the updated binary classification decision model.

[0032] When the loss curve converges, the drill-down analysis decision model that has been trained is obtained.

[0033] To address the above problems, the present invention also provides an adaptive table-granularity data generation device, the device comprising:

[0034] The data dimension initialization module is used to obtain raw data from a pre-built business system cluster, sort the raw data by indicator importance to obtain a business important dimension order table, and obtain the initial dimension according to the business important dimension order table;

[0035] The table granularity divergence analysis module is used to perform table granularity analysis on the original data according to the initialization dimension, and to obtain the divergence of the original data under the initialization dimension.

[0036] The drill-down analysis and judgment module is used to determine whether to perform drill-down analysis on the original data based on the divergence situation using a pre-trained drill-down analysis decision model.

[0037] The loop traversal module is used to add dimensions to the initial dimensions according to the preset business importance dimension order table when the judgment result is that drill-down analysis of the original data is required, to obtain the updated dimensions, and return to the operation of the table granularity divergence analysis module to perform table granularity analysis on the original data according to the updated dimensions.

[0038] The data generation module is used to configure the initial dimension as an adaptive table granularity when the judgment result is that drill-down analysis of the original data is not required, and to perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0039] Optionally, the step of performing table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension includes:

[0040] Obtain the first part of the original data with a total number of statistical records greater than 1 under the initialization dimension, and obtain the number of records and creation time of the first part of the original data;

[0041] Obtain the second part of the original data where the number of summary statistics entries under the initialization dimension is equal to 1, and obtain the number of entries and creation time of the second part of the original data;

[0042] Based on the number of records and creation time of the first part of the original data, a histogram is constructed to obtain the data distribution; and based on the number of records and creation time of the second part of the original data, a line chart is constructed to obtain the data change trend.

[0043] The data distribution and the data change trend are combined and output to obtain the divergence of the original data under the initialization dimension.

[0044] Optionally, the step of using a pre-trained drill-down analysis decision model to determine whether to perform drill-down analysis on the original data based on the divergence situation includes:

[0045] The divergence features are extracted from the divergence situation using a pre-trained drill-down analysis decision model to obtain a divergence feature sequence;

[0046] Based on the initialization dimension, feature engineering is performed on the original data to extract features, resulting in a business feature sequence;

[0047] Based on the business feature sequence, determine the business problem complexity of the initialization dimension, and based on the divergent feature sequence, determine whether the dispersion condition satisfies the business problem complexity.

[0048] When the dispersion situation cannot meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

[0049] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0050] At least one processor; and,

[0051] A memory communicatively connected to the at least one processor; wherein,

[0052] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the adaptive table-granularity data generation method described above.

[0053] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one computer program, which is executed by a processor in an electronic device to implement the adaptive table-granularity data generation method described above.

[0054] This invention first selects a suitable business system cluster based on preset business requirements. Then, it performs principal component analysis on the raw data in the business system cluster to identify key factors for solving business needs, obtaining a business importance dimension order list. Next, based on this order list, it extracts several dimension construction points to initialize dimensions and performs table granularity analysis to obtain divergence. This divergence includes the data distribution and trends. Analysts typically understand the distribution and trends of data across different dimensions to make more accurate decisions. This invention employs a drill-down analysis decision model to decide whether to perform data drill-down, resulting in faster and more accurate results. The drill-down analysis decision model is first constructed based on a binary classification problem of whether to perform drill-down, then selects appropriate input parameters through feature engineering, and finally trains using a cross-entropy loss algorithm, exhibiting high computational efficiency and accuracy. Therefore, this invention provides an adaptive table granularity data generation method, apparatus, device, and storage medium that can generate more scientifically sound and reasonable data for analysis in financial big data exploration scenarios by adaptively identifying table granularity suitable for business problems. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating an adaptive table-granularity data generation method according to an embodiment of the present invention.

[0056] Figure 2 This is a detailed flowchart illustrating one step in an adaptive table-granularity data generation method according to an embodiment of the present invention.

[0057] Figure 3 This is a detailed flowchart illustrating one step in an adaptive table-granularity data generation method according to an embodiment of the present invention.

[0058] Figure 4 This is a functional block diagram of an adaptive table-granularity data generation device provided in an embodiment of the present invention;

[0059] Figure 5 This is a schematic diagram of the structure of an electronic device that implements the adaptive table granularity data generation method according to an embodiment of the present invention.

[0060] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0061] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0062] This application provides an adaptive table-level data generation method. In this application, the executing entity of the adaptive table-level data generation method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the adaptive table-level data generation method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0063] Reference Figure 1 The diagram shown is a flowchart illustrating an adaptive table-granularity data generation method according to an embodiment of the present invention. In this embodiment, the adaptive table-granularity data generation method includes steps S1 to S4:

[0064] S1. Obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business importance dimension order table, and obtain the initial dimension according to the business importance dimension order table.

[0065] In this embodiment of the invention, the business system cluster is selected according to the specific problems that need to be solved in the business. For example, when mining insurance policies, the original data from the source can be obtained from systems such as reinsurance system, underwriting system, and claims investigation system.

[0066] Furthermore, the ranking of indicator importance refers to prioritizing indicators in the original data that are more relevant to solving the problem, such as characteristics or attributes of the data, which may include time, location, product, etc.

[0067] In detail, in this embodiment of the invention, obtaining raw data from a pre-built business system cluster and sorting the raw data by indicator importance to obtain a business importance dimension order table includes:

[0068] Based on the preset business requirements, a cluster of business systems is selected from the pre-built enterprise main system;

[0069] The raw data is obtained from the business system cluster, and principal component analysis is performed on the raw data to obtain a set of principal component fields.

[0070] The principal component field set is fine-tuned according to the pre-built expert configuration labels to obtain a business importance dimension order table.

[0071] Principal Component Analysis (PCA) is a commonly used dimensionality reduction technique and data analysis method. It reduces the dimensionality of the data and retains the most important information by projecting the original data onto a new coordinate system and finding the projection direction that can preserve the variance of the original data to the greatest extent.

[0072] Furthermore, in this embodiment of the invention, the expert configuration label refers to the label that the analysis expert assigns weights to each important indicator based on business experience.

[0073] In this embodiment of the invention, the principal component analysis results can be fine-tuned by configuring the expert tags to obtain a more suitable order table of business importance dimensions that conforms to the characteristics of the enterprise and the business scenario. Then, the most important ones, such as three fields, are extracted as the initial dimensions.

[0074] S2. Perform table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension.

[0075] The divergence refers to the distribution and trend of the data.

[0076] For details, please refer to the following: Figure 2 As shown in this embodiment of the invention, the step of performing table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension includes steps S21 to S24:

[0077] S21. Obtain the first part of the original data with a total number of statistical records greater than 1 under the initialization dimension, and obtain the number of records and creation time of the first part of the original data.

[0078] S22. Obtain the second part of the original data where the number of summary statistics entries under the initialization dimension is equal to 1, and obtain the number of entries and creation time of the second part of the original data.

[0079] S23. Based on the number of records and creation time of the first part of the original data, construct a histogram to obtain the data distribution; and based on the number of records and creation time of the second part of the original data, construct a line chart to obtain the data change trend.

[0080] S24. Combine and output the data distribution and the data change trend to obtain the divergence of the original data under the initialization dimension.

[0081] The divergence index described in this embodiment of the invention includes two parts of data: first, the number of data entries with a total number of entries greater than 1 and their creation time under this dimension, i.e., data with multiple values ​​of the same dimension; second, the number of data entries with a total number of entries equal to 1 and their creation time under this dimension, i.e., data with only one value of this dimension.

[0082] By analyzing these indicators, this invention can understand the distribution and trends of data in different dimensions, thereby making more accurate decisions. The total number of statistical entries refers to the result of statistical analysis of data in that dimension according to a certain aggregation method (such as summation, counting, etc.).

[0083] This invention can understand the distribution of data across different values ​​by statistically analyzing the frequency or proportion of the original data under the initialization dimension. The distribution of data can be displayed using visualization methods such as histograms, pie charts, or bar charts. Furthermore, by observing the changing trend of the original data under the initialization dimension, the changes of the data at different time points or under different conditions can be understood. The changing trend of the data can be displayed using visualization methods such as line charts and trend charts.

[0084] S3. Using a pre-trained drill-down analysis decision model, determine whether to perform drill-down analysis on the original data based on the divergence situation.

[0085] In this embodiment of the invention, the drill-down analysis decision model is a binary classification model based on a neural network, used to determine whether it is necessary to add dimensions for drill-down analysis based on the original data and the divergence of the original data under the initial dimension.

[0086] For details, please refer to the following: Figure 3 As shown in this embodiment of the invention, the step of using a pre-trained drill-down analysis decision model to determine whether to perform drill-down analysis on the original data based on the divergence situation includes steps S31 to S34:

[0087] S31. Using a pre-trained drill-down analysis decision model, the divergence features of the divergence situation are extracted to obtain a divergence feature sequence;

[0088] S32. Based on the initialization dimension, perform feature engineering and feature extraction on the original data to obtain a business feature sequence;

[0089] S33. Based on the business feature sequence, determine the business problem complexity of the initialization dimension, and based on the divergent feature sequence, determine whether the dispersion condition satisfies the business problem complexity.

[0090] S34. When the dispersion situation cannot meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

[0091] In this embodiment of the invention, determining whether further drill-down analysis of the data is needed requires considering the importance of the data to the current problem or objective. The initial dimensions described in this embodiment are all relatively important fields, therefore, feature extraction can be performed directly to obtain a business feature sequence. Furthermore, in this embodiment, determining whether to perform drill-down analysis also requires considering the complexity and depth of the problem. If the problem is complex and requires more comprehensive data support and analysis, further in-depth analysis may be necessary. Therefore, this invention needs to determine whether the dispersion situation meets the complexity of the business problem. When the dispersion situation does not meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

[0092] In detail, in this embodiment of the invention, before utilizing the pre-trained drill-down analysis decision model, the method further includes:

[0093] Obtain the pre-built initial binary classification decision model and the pre-processed business scenario data set;

[0094] Feature engineering is performed on the business scenario data set to extract training sample sets;

[0095] Target samples are extracted sequentially from the training sample set, and the initial binary classification decision model is used to perform drill-down analysis on the target samples to obtain prediction results.

[0096] Using the cross-entropy loss algorithm, the loss value between the true label corresponding to the target sample and the prediction result is calculated, and the loss value is minimized to obtain the model parameters when the loss value is minimized. Based on the model parameters, the initial binary classification decision model is updated by reverse network to obtain the updated binary classification decision model.

[0097] Record the loss values, obtain the loss value curve, and determine the convergence of the loss value curve;

[0098] When the loss curve fails to converge, return to the above operation of extracting target samples from the training sample set in sequence, and iteratively optimize the updated binary classification decision model.

[0099] When the loss curve converges, the drill-down analysis decision model that has been trained is obtained.

[0100] This invention, based on business requirements, clarifies the problem to be solved and determines the model's objective: whether to further increase the dimensions in the initialization dimension for drill-down analysis. Therefore, this invention constructs a binary classification decision model. Then, necessary data, such as a business scenario dataset, is collected and preprocessed and cleaned to ensure data quality and consistency. Next, based on the problem's requirements and data characteristics, appropriate features are selected and feature engineering is performed, such as feature extraction, feature transformation, and feature combination, to obtain a training sample set. Finally, the model is trained using the cross-entropy loss algorithm to obtain the trained drill-down analysis decision model.

[0101] When the determination result indicates that drill-down analysis of the original data is required, S4, add dimensions to the initial dimensions according to the preset business importance dimension order table to obtain the updated dimensions, and return to the operation of step B, and perform table granular analysis on the original data according to the updated dimensions.

[0102] In this embodiment of the invention, when the determination result indicates that drill-down analysis of the original data is required, a dimension needs to be added to the initial dimension to obtain an updated dimension. Then, the table granularity analysis in step S2 is performed again to check the data distribution and data change trend of the original data under the updated dimension, and to determine whether further drill-down is required.

[0103] When the determination result is that drill-down analysis of the original data is not required, S5, configure the initial dimension as the adaptive table granularity, and perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0104] In this solution, the drill-down depth can be one, two, or even multiple layers. If the initialization or update dimension is determined to be at the table granularity, the data exploration platform can provide data cleaning options based on the table granularity, such as discarding, archiving, entering a cold partition, formatting, or creating a new table. This allows for data cleaning and processing to obtain data with adaptive table granularity.

[0105] This invention first selects a suitable business system cluster based on preset business requirements. Then, it performs principal component analysis on the raw data in the business system cluster to identify key factors for solving business needs, obtaining a business importance dimension order list. Next, based on this order list, it extracts several dimension construction points to initialize dimensions and performs table granularity analysis to obtain divergence. This divergence includes the distribution and trends of the data. Analysts typically understand the distribution and trends of data across different dimensions to make more accurate decisions. This invention employs a drill-down analysis decision model to decide whether to perform data drill-down, resulting in faster and more accurate results. The drill-down analysis decision model is first constructed based on a binary classification problem of whether to perform drill-down, then selects appropriate input parameters through feature engineering, and finally trains using a cross-entropy loss algorithm, exhibiting high computational efficiency and accuracy. Therefore, this invention provides an adaptive table granularity data generation method that can generate more scientifically sound and reasonable data for analysis in financial big data exploration scenarios by adaptively identifying table granularity suitable for business problems.

[0106] like Figure 4 The diagram shown is a functional block diagram of an adaptive table-granularity data generation device provided in an embodiment of the present invention.

[0107] The adaptive table-granularity data generation device 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the adaptive table-granularity data generation device 100 may include a data dimension initialization module 101, a table granularity divergence analysis module 102, a drill-down analysis and judgment module 103, a loop traversal module 104, and a data generation module 105. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0108] In this embodiment, the functions of each module / unit are as follows:

[0109] The data dimension initialization module 101 is used to obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business important dimension order table, and obtain the initial dimension according to the business important dimension order table.

[0110] The table granularity divergence analysis module 102 is used to perform table granularity analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension.

[0111] The drill-down analysis and judgment module 103 is used to determine whether to perform drill-down analysis on the original data based on the divergence situation using a pre-trained drill-down analysis decision model.

[0112] The loop traversal module 104 is used to add dimensions to the initial dimensions according to the preset business importance dimension order table when the judgment result is that drill-down analysis of the original data is required, to obtain the updated dimensions, and return to the operation of the table granularity divergence analysis module to perform table granularity analysis on the original data according to the updated dimensions.

[0113] The data generation module 105 is used to configure the initial dimension as an adaptive table granularity when the judgment result is that drill-down analysis of the original data is not required, and to perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0114] In detail, the modules in the adaptive table-granularity data generation device 100 described in this application embodiment adopt the same usage as described above. Figures 1 to 3 The adaptive table granularity data generation method described herein uses the same technical means and can produce the same technical effect, so it will not be elaborated here.

[0115] like Figure 5 The diagram shown is a structural schematic of an electronic device 1 that implements an adaptive table granularity data generation method according to an embodiment of the present invention.

[0116] The electronic device 1 may include a processor 10, a memory 11, a communication bus 12, and a communication interface 13. It may also include a computer program stored in the memory 11 and capable of running on the processor 10, such as an adaptive table-granularity data generation program.

[0117] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device 1, connecting various components of the electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., executing adaptive table-granularity data generation programs) and calls data stored in the memory 11 to perform various functions of the electronic device and process data.

[0118] The memory 11 includes at least one type of readable storage medium, including flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of an electronic device, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device, such as a plug-in portable hard drive, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc. Furthermore, the memory 11 can include both internal and external storage units of the electronic device. The memory 11 can be used not only to store application software and various types of data installed on the electronic device, such as the code of an adaptive table-granularity data generation program, but also to temporarily store data that has been output or will be output.

[0119] The communication bus 12 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. The bus is configured to enable communication between the memory 11 and at least one processor 10, etc.

[0120] The communication interface 13 is used for communication between the electronic device 1 and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, Bluetooth interface, etc.), typically used to establish communication connections between the electronic device and other electronic devices. The user interface may be a display, an input unit (such as a keyboard), or, optionally, a standard wired or wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device and to display a visual user interface.

[0121] Figure 5 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 5The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0122] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0123] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0124] The adaptive table-granularity data generation program stored in the memory 11 of the electronic device 1 is a combination of multiple instructions, which, when run in the processor 10, can achieve the following:

[0125] Step A: Obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business importance dimension order table, and obtain the initial dimension according to the business importance dimension order table;

[0126] Step B: Perform table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension;

[0127] Step C: Using a pre-trained drill-down analysis decision model, determine whether to perform drill-down analysis on the original data based on the divergence situation;

[0128] When the determination result indicates that drill-down analysis of the original data is required, step D: add dimensions to the initial dimensions according to the preset business importance dimension order table to obtain the updated dimensions, and return to the operation of step B, and perform table granular analysis on the original data according to the updated dimensions;

[0129] When the determination result is that drill-down analysis of the original data is not required, step E: configure the initial dimension as the adaptive table granularity, and perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0130] Specifically, the specific implementation method of the processor 10 for the above instructions can be referred to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, and will not be repeated here.

[0131] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0132] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following:

[0133] Step A: Obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business importance dimension order table, and obtain the initial dimension according to the business importance dimension order table;

[0134] Step B: Perform table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension;

[0135] Step C: Using a pre-trained drill-down analysis decision model, determine whether to perform drill-down analysis on the original data based on the divergence situation;

[0136] When the determination result indicates that drill-down analysis of the original data is required, step D: add dimensions to the initial dimensions according to the preset business importance dimension order table to obtain the updated dimensions, and return to the operation of step B, and perform table granular analysis on the original data according to the updated dimensions;

[0137] When the determination result is that drill-down analysis of the original data is not required, step E: configure the initial dimension as the adaptive table granularity, and perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity.

[0138] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0139] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0140] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0141] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0142] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0143] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0144] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0145] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices recited in a system claim may also be implemented by a single unit or device through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A data generation method with adaptive table granularity, characterized in that, The method includes: Step A: Obtain raw data from the pre-built business system cluster, sort the raw data by indicator importance to obtain a business importance dimension order table, and obtain the initial dimension according to the business importance dimension order table; Step B: Perform table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension. The divergence is the distribution and trend of the data. Step C: Using a pre-trained drill-down analysis decision model, determine whether to perform drill-down analysis on the original data based on the divergence situation; When the determination result indicates that drill-down analysis of the original data is required, step D: add dimensions to the initial dimensions according to the preset business importance dimension order table to obtain the updated dimensions, and return to the operation of step B, and perform table granular analysis on the original data according to the updated dimensions; When the determination result is that drill-down analysis of the original data is not required, step E: configure the initial dimension as the adaptive table granularity, and perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity; The step of performing table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension includes: Obtain the first part of the original data with a total number of statistical records greater than 1 under the initialization dimension, and obtain the number of records and creation time of the first part of the original data; Obtain the second part of the original data where the number of summary statistics entries under the initialization dimension is equal to 1, and obtain the number of entries and creation time of the second part of the original data; Based on the number of records and creation time of the first part of the original data, a histogram is constructed to obtain the data distribution; and based on the number of records and creation time of the second part of the original data, a line chart is constructed to obtain the data change trend. The data distribution and the data change trend are combined and output to obtain the divergence of the original data under the initialization dimension.

2. The adaptive table-granularity data generation method as described in claim 1, characterized in that, The step of using a pre-trained drill-down analysis decision model to determine whether to perform drill-down analysis on the original data based on the divergence situation includes: The divergence features are extracted from the divergence situation using a pre-trained drill-down analysis decision model to obtain a divergence feature sequence; Based on the initialization dimension, feature engineering is performed on the original data to extract features, resulting in a business feature sequence; Based on the business feature sequence, determine the business problem complexity of the initialization dimension, and based on the divergence feature sequence, determine whether the divergence condition satisfies the business problem complexity. When the divergence condition cannot meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

3. The adaptive table-granularity data generation method as described in claim 1, characterized in that, The process of obtaining raw data from a pre-built business system cluster and sorting the raw data by indicator importance to obtain a business importance dimension order table includes: Based on the preset business requirements, a cluster of business systems is selected from the pre-built enterprise main system; The raw data is obtained from the business system cluster, and principal component analysis is performed on the raw data to obtain a set of principal component fields. The principal component field set is fine-tuned according to the pre-built expert configuration labels to obtain a business importance dimension order table.

4. The adaptive table-granularity data generation method as described in claim 1, characterized in that, Prior to utilizing the pre-trained drill-down analysis decision model, the method further includes: Obtain the pre-built initial binary classification decision model and the pre-processed business scenario data set; Feature engineering is performed on the business scenario data set to extract training sample sets; Target samples are extracted sequentially from the training sample set, and the initial binary classification decision model is used to perform drill-down analysis on the target samples to obtain prediction results. Using the cross-entropy loss algorithm, the loss value between the true label corresponding to the target sample and the prediction result is calculated, and the loss value is minimized to obtain the model parameters when the loss value is minimized. Based on the model parameters, the initial binary classification decision model is updated by reverse network to obtain the updated binary classification decision model. Record the loss values, obtain the loss value curve, and determine the convergence of the loss value curve; When the loss curve fails to converge, return to the above operation of extracting target samples from the training sample set in sequence, and iteratively optimize the updated binary classification decision model. When the loss curve converges, the drill-down analysis decision model that has been trained is obtained.

5. A data generation device with adaptive table granularity, characterized in that, The device includes: The data dimension initialization module is used to obtain raw data from a pre-built business system cluster, sort the raw data by indicator importance to obtain a business important dimension order table, and obtain the initial dimension according to the business important dimension order table; The table granularity divergence analysis module is used to perform table granularity analysis on the original data according to the initialization dimension, and to obtain the divergence of the original data under the initialization dimension. The divergence is the distribution and trend of the data. The drill-down analysis and judgment module is used to determine whether to perform drill-down analysis on the original data based on the divergence situation using a pre-trained drill-down analysis decision model. The loop traversal module is used to add dimensions to the initial dimensions according to the preset business importance dimension order table when the judgment result is that drill-down analysis of the original data is required, to obtain the updated dimensions, and return to the operation of the table granularity divergence analysis module to perform table granularity analysis on the original data according to the updated dimensions. The data generation module is used to configure the initial dimension as an adaptive table granularity when the judgment result is that drill-down analysis of the original data is not required, and to perform data cleaning on the original data according to the adaptive table granularity to generate data with adaptive table granularity. The step of performing table-level analysis on the original data according to the initialization dimension to obtain the divergence of the original data under the initialization dimension includes: Obtain the first part of the original data with a total number of statistical records greater than 1 under the initialization dimension, and obtain the number of records and creation time of the first part of the original data; Obtain the second part of the original data where the number of summary statistics entries under the initialization dimension is equal to 1, and obtain the number of entries and creation time of the second part of the original data; Based on the number of records and creation time of the first part of the original data, a histogram is constructed to obtain the data distribution; and based on the number of records and creation time of the second part of the original data, a line chart is constructed to obtain the data change trend. The data distribution and the data change trend are combined and output to obtain the divergence of the original data under the initialization dimension.

6. The adaptive table-granularity data generation device as described in claim 5, characterized in that, The step of using a pre-trained drill-down analysis decision model to determine whether to perform drill-down analysis on the original data based on the divergence situation includes: The divergence features are extracted from the divergence situation using a pre-trained drill-down analysis decision model to obtain a divergence feature sequence; Based on the initialization dimension, feature engineering is performed on the original data to extract features, resulting in a business feature sequence; Based on the business feature sequence, determine the business problem complexity of the initialization dimension, and based on the divergence feature sequence, determine whether the divergence condition satisfies the business problem complexity. When the divergence condition cannot meet the complexity of the business problem, it is determined that drill-down analysis of the original data is required.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the adaptive table-granularity data generation method as described in any one of claims 1 to 4.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the adaptive table-granularity data generation method as described in any one of claims 1 to 4.

Images