Data analysis method and system based on artificial intelligence

Abstract

The invention discloses a data analysis method and system based on artificial intelligence. The data analysis method based on the artificial intelligence comprises the following steps that data collection is carried out; classifying according to the collected data, and determining a data analysis set; carrying out distributed storage on the data analysis set; and determining a data analysis strategy after distributed storage, and analyzing the data analysis set according to the data analysis strategy. According to the method and the device, the data can be distinguished and divided in the data analysis process, so that different analysis processing is performed for different data, the analysis accuracy is improved, and the analysis execution time is shortened.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular, to an artificial intelligence-based data analysis method and system thereof. Background Art

[0002] With the development of digital technology, data analysis is of vital importance, and data analysis based on artificial intelligence is the frontier direction in the current data analysis field. It combines artificial intelligence technology with traditional data analysis methods to provide a more powerful means for extracting valuable information from massive data. Artificial intelligence can process large-scale data. Artificial intelligence algorithms can efficiently process large-scale data sets and dig out patterns and laws hidden in the data. At the same time, data analysis can be automated, reducing manual intervention and improving analysis efficiency. Through trained models, new data can be quickly analyzed and predicted to provide timely support for decision-making. And complex relationships can be discovered in a timely manner. Artificial intelligence algorithms can discover complex nonlinear relationships in data, while traditional data analysis methods can often only process linear relationships. This enables data analysis based on artificial intelligence to understand data more deeply and provide more accurate predictions and decision-making suggestions. However, in this way of data analysis, there are still inaccurate analysis caused by data quality problems, because inaccurate, incomplete or inconsistent data will affect the reliability of the analysis results. Even if artificial intelligence can process large-scale data sets at high speed, it will process inaccurate and incomplete data together with accurate data for analysis, which will increase the analysis time and reduce the accuracy of the analysis.

[0003] Therefore, how to provide a method that can distinguish different data and perform different analyses has become an urgent problem to be solved in this field. Summary of the invention

[0004] The present application proposes an artificial intelligence-based data analysis method, comprising the following steps: collecting data; classifying the collected data to determine a data analysis set; performing distributed storage on the data analysis set; determining a data analysis strategy after distributed storage, and analyzing the data analysis set according to the data analysis strategy.

[0005] As above, wherein, classifying the collected data and determining the data analysis set includes the following sub-steps: determining whether the collected data has missing data; classifying the collected data according to the missing results and determining the data analysis set.

[0006] As above, if the collected data is missing, the missing data will be used to form a non-high-quality data analysis set; if the collected data is not missing, the non-missing data will be used to form a high-quality data analysis set.

[0007] As mentioned above, determining whether data is missing includes performing data verification, performing data consistency check, and performing data integrity constraint check.

[0008] As mentioned above, the distributed storage of the data analysis sets includes interrupting the storage of high-quality data analysis sets and continuously storing low-quality data analysis sets.

[0009] A data analysis system based on artificial intelligence, specifically comprising: a collection unit, an analysis set determination unit, a storage unit and an analysis strategy determination unit; the collection unit is used to collect data; the analysis set determination unit is used to classify the collected data and determine the data analysis set; the storage unit is used to perform distributed storage of the data analysis set; the analysis strategy determination unit is used to determine the data analysis strategy and analyze the data analysis set according to the data analysis strategy.

[0010] As above, the analysis set determination unit classifies the collected data, and determining the data analysis set includes the following sub-steps: determining whether the collected data has missing data; classifying the collected data according to the missing results, and determining the data analysis set.

[0011] As above, if the collected data is missing, the analysis set determination unit will form a non-high-quality data analysis set with the missing data; if the collected data is not missing, the non-missing data will be formed into a high-quality data analysis set.

[0012] As mentioned above, the analysis set determination unit determines whether data is missing, including performing data verification, data consistency check and data integrity constraint check.

[0013] As described above, the storage unit performs distributed storage on the data analysis sets, including interrupting the storage of high-quality data analysis sets and continuously storing low-quality data analysis sets.

[0014] This application has the following beneficial effects:

[0015] The present application can distinguish and divide data during the data analysis process, thereby performing different analysis and processing on different data, improving the accuracy of the analysis and speeding up the execution time of the analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] 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 the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 is a flow chart of a data analysis method based on artificial intelligence provided in an embodiment of the present application;

[0018] Figure 2 It is a schematic diagram of the internal structure of an artificial intelligence-based data analysis system provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0020] Embodiment 1

[0021] like Figure 1 As shown, this embodiment provides a data analysis method based on artificial intelligence, which specifically includes the following steps:

[0022] Step S1: data collection;

[0023] Data can be collected in many aspects, and the collected data is generally large in volume and needs to be analyzed.

[0024] Step S2: Classify the collected data and determine the data analysis set;

[0025] Since the volume of collected data is large, if data analysis is performed directly, the current computer computing power cannot cope with the demand for fast processing response. Therefore, it is necessary to classify and screen the acquired data, and perform different analysis methods based on different classification and screening results.

[0026] According to the collected data, the classification and screening are carried out to determine the data analysis set, which includes the following sub-steps:

[0027] Step S21: Determine whether there is any missing in the collected data.

[0028] Obtaining accurate and complete data is the basis for making wise decisions. If the data is incomplete or erroneous, it may lead to wrong decisions, thus affecting the development and success of the business. Therefore, this embodiment needs to determine whether the collected data is missing.

[0029] The following methods can be used to determine whether data is missing:

[0030] Data validation: Checks whether the format, type, and range of the data are as expected. For example, check whether the date field is in the correct date format and whether the number field is within a reasonable range.

[0031] Data validation can be performed using regular expressions, data validation tools, or data type checking functions in programming languages. If the result of data validation does not meet expectations, it is considered that the collected data is missing, otherwise it is considered that the collected data is not missing.

[0032] Consistency check: Compare the data in different data sources or different parts of the same data source to see if they are consistent. For example, check if the customer information in the database is consistent with the information in the customer service system.

[0033] If the data are inconsistent, it is considered that the collected data are missing. If the data are consistent, it is considered that the collected data are not missing.

[0034] Integrity constraint check: Create the database in advance. The database usually has integrity constraints, such as primary key constraints, foreign key constraints, unique constraints, etc. Checking whether these constraints are met can ensure the integrity of the data. If the above constraints are not met, it is considered that the collected data is missing, otherwise it is considered that the collected data is not missing.

[0035] Step S22: Classify the collected data according to the missing results and determine the data analysis set.

[0036] The data analysis set includes a high-quality data analysis set and a non-high-quality data analysis set. The data considered to be missing in the above steps are grouped together to form a non-high-quality data analysis set, and the data considered to be not missing are grouped together to form a high-quality data analysis set.

[0037] Step S3: Distributed storage of the data analysis set;

[0038] In this embodiment, different storage methods are used for different data analysis sets. The distributed storage method proposed in this embodiment includes continuous storage and interrupt storage. After distributed storage, the corresponding data can be read out at the storage location to analyze the corresponding data.

[0039] The storage location may be a virtual storage node, and corresponding data is stored in different nodes, and corresponding stored data is read from the corresponding node.

[0040] Continuous storage is repeated storage, and interrupt storage is interrupted after the current storage is performed once, and the next storage is performed when the next instruction arrives.

[0041] For high-quality data analysis sets, since their integrity can be guaranteed, the data therein are relatively easy to process. Therefore, it is expected that the processing and analysis of the high-quality data analysis sets can be performed in a shorter time, so the high-quality data analysis sets are stored interrupted.

[0042] For non-high-quality data analysis sets, since they may be missing, the data processing therein is relatively difficult, and there may be problems such as anomalies when processing data. Therefore, this embodiment stores them in a continuous storage manner. If data errors occur, the stored data is retrieved for easy reprocessing.

[0043] It can be understood that continuous storage is to repeatedly store non-quality data over and over again. After each storage, new non-quality data is stored again. Theoretically, the non-quality data stored this time and last time should be the same data. If it is an interrupt storage mode, after reading the data at the corresponding storage location, the data at the location is cleared accordingly until the next storage instruction is issued to store the corresponding data.

[0044] Step S4: After distributed storage, determine the data analysis strategy, and analyze the data analysis set according to the data analysis strategy.

[0045] Determining the data analysis strategy specifically includes the following sub-steps:

[0046] Step S41: Determine the analysis strategy of the high-quality data analysis data set, and analyze the high-quality data in the high-quality data analysis set.

[0047] The strategy for analyzing high-quality data sets is to read the corresponding high-quality data from the storage location and directly analyze it.

[0048] The analysis method may refer to the regression analysis, correlation analysis and other methods in the prior art, and the specific methods are not described in detail here.

[0049] Step S42: Determine a non-high-quality data set analysis strategy, and perform analysis on the non-high-quality data in the non-high-quality data analysis set.

[0050] The non-high-quality data set analysis strategy is: read the corresponding non-high-quality data from the storage location, perform anomaly detection on the non-high-quality data, analyze it if no anomaly is detected, and return to step S3 if an anomaly is detected, and re-acquire the stored non-high-quality data after distributed storage.

[0051] Step S42 specifically includes the following sub-steps:

[0052] Step S421: Scan the non-high-quality data to determine data points of the non-high-quality data.

[0053] The non-high-quality data X = {x1, x2, ...xi, ...xN} includes multiple data points, N is the size of the non-high-quality data, xi is a single data point in the non-high-quality data, and each data point contains multiple elements, denoted as xi1, xi2, ...xij, ...xiD, where each data point xi is a D-tuple of category A = (A1, A2, ...AD), A1, A2, ...AD represent different attributes of the elements, and each element can have the same or different attributes. Category A is only a general term and does not mean that it belongs to category A.

[0054] Step S422: divide the data points of the non-high-quality data to determine whether there are abnormal data points.

[0055] The point division is to divide different data points into point sets. If there are data points that do not belong to the point set, they are regarded as abnormal data points.

[0056] The point division specifically includes the following sub-steps:

[0057] Step S4221: Determine the initial center point and the initial point set.

[0058] A data point among multiple data points is randomly selected as an initial center point, and an initial point set is determined according to the initial center point.

[0059] The point set is in the r neighborhood, with the initial center point as the center. A circle with a radius of . Each data point represents a category. For example, data point xi is selected as the cluster center of category A.

[0060] Step S4222: Update the initial point set and determine abnormal data points.

[0061] Continue to scan the data points other than the initial center point in the non-high-quality data. If the distance R(x A ,x i ) is less than or equal to If is the radius, the data point is put into the point set, completing a point set update, and the point set is continued to be updated by distance comparison.

[0062]

[0063] Among them, x A represents the initial center of category A, x i Represents a data point selected for distance calculation with the initial center, where x Aj Represents the initial center x of category A A Any element contained in x ijRepresents data point x i Any element contained, n represents the number of elements contained in the data points in the selected category A, j is a natural number, (x Aj ,AD) represents the initial center x of category A A The attribute contained is the number of AD elements, (x ij ,AD) data point x of category A i The attribute contained is the number of AD elements.

[0064] Furthermore, if the distance between a data point and the initial center point is greater than If the radius is the radius, then the point needs to be used as another initial center point, the initial center point is defined as a new center point, and a new initial point set is determined based on the new center point. The determination method of the new initial point set refers to step S4221, which is not described in detail here. The distance between the remaining data points and the new center point is compared to update the new initial point set. The distance between the remaining data points and the new center point refers to the above formula, and the specific method is not described in detail here.

[0065] The update of one or more point sets is completed according to the above method, and each point set contains corresponding data points. If all the data points are included in the corresponding point set, it is considered that there are no abnormal data points, and non-quality data analysis is performed. The analysis method can refer to the regression analysis, correlation analysis and other methods in the prior art, and the specific method is not repeated here.

[0066] If there is a data point that is not included in one or more point sets, the data point is considered to be an abnormal data point. If an abnormal data point appears, return to step S3 and re-read the stored non-high-quality data.

[0067] Since there may be multiple reads of non-quality data from multiple storage locations, it is also necessary to determine the access limit Q. If the access limit Q reaches the upper limit, the reading of non-quality data from one or more storage locations is stopped, and the non-quality data is stored in other storage locations.

[0068] The access quota Q is expressed as:

[0069]

[0070] where t p represents the number of times the non-quality data is accessed in the storage location p, P represents the total number of storage locations storing non-quality data, and v p represents the routing cost required to read non-quality data, ω P Represents the storage coefficient of the storage location p where non-quality data is stored. Generally speaking, the more times non-quality data is stored, the smaller the coefficient is, and vice versa. PThe specific value is set in advance by the staff and is not limited here.

[0071] Embodiment 2

[0072] like Figure 2 As shown, this embodiment provides an artificial intelligence-based data analysis system, which specifically includes: a collection unit 210, an analysis set determination unit 220, a storage unit 230 and an analysis strategy determination unit 240.

[0073] A collection unit 210, configured to collect data;

[0074] Data can be collected in many aspects, and the collected data is generally large in volume and needs to be analyzed.

[0075] An analysis set determination unit 220, configured to classify the collected data and determine a data analysis set;

[0076] The analysis set determination unit 220 includes the following submodules:

[0077] The missing data judgment module is used to judge whether the collected data is missing.

[0078] The following methods can be used to determine whether data is missing:

[0079] Data validation: Checks whether the format, type, and range of the data are as expected. For example, check whether the date field is in the correct date format and whether the number field is within a reasonable range.

[0080] Data validation can be performed using regular expressions, data validation tools, or data type checking functions in programming languages. If the result of data validation does not meet expectations, it is considered that the collected data is missing, otherwise it is considered that the collected data is not missing.

[0081] Consistency check: Compare the data in different data sources or different parts of the same data source to see if they are consistent. For example, check if the customer information in the database is consistent with the information in the customer service system.

[0082] If the data are inconsistent, it is considered that the collected data are missing. If the data are consistent, it is considered that the collected data are not missing.

[0083] Integrity constraint check: Create the database in advance. The database usually has integrity constraints, such as primary key constraints, foreign key constraints, unique constraints, etc. Checking whether these constraints are met can ensure the integrity of the data. If the above constraints are not met, it is considered that the collected data is missing, otherwise it is considered that the collected data is not missing.

[0084] The data analysis set determination module is used to classify the collected data according to the missing results and determine the data analysis set.

[0085] The data analysis set includes a high-quality data analysis set and a non-high-quality data analysis set. The data considered to be missing in the above steps are grouped together to form a non-high-quality data analysis set, and the data considered to be not missing are grouped together to form a high-quality data analysis set.

[0086] Storage unit 230, used for distributed storage of data analysis sets;

[0087] In this embodiment, different storage methods are used for different data analysis sets. The distributed storage method proposed in this embodiment includes continuous storage and interrupt storage. After distributed storage, the corresponding data can be read out at the storage location to analyze the corresponding data.

[0088] Continuous storage is repeated storage, and interrupt storage is interrupted after the current storage is performed once, and the next storage is performed when the next instruction arrives.

[0089] For high-quality data analysis sets, since their integrity can be guaranteed, the data therein are relatively easy to process. Therefore, it is expected that the processing and analysis of the high-quality data analysis sets can be performed in a shorter time, so the high-quality data analysis sets are stored interrupted.

[0090] For non-high-quality data analysis sets, since they may be missing, the data processing therein is relatively difficult, and there may be problems such as anomalies when processing data. Therefore, this embodiment stores them in a continuous storage manner. If data errors occur, the stored data is retrieved for easy reprocessing.

[0091] It can be understood that continuous storage is the repeated storage of non-high-quality data over and over again. After obtaining non-high-quality data after each storage, new non-high-quality data is stored again. In theory, the non-high-quality data stored this time and last time should be the same data.

[0092] The analysis strategy determination unit 240 is used to determine the data analysis strategy and analyze the data analysis set according to the data analysis strategy.

[0093] Determine the data analysis strategy specifically including the following submodules:

[0094] The module for determining the analysis strategy of the high-quality data analysis data set is used to determine the analysis strategy of the high-quality data analysis data set and analyze the high-quality data in the high-quality data analysis set. The analysis strategy of the high-quality data set is: read the corresponding high-quality data from the storage location and directly analyze it. The analysis method can refer to the regression analysis, correlation analysis and other methods in the prior art, and the specific method content is not repeated here.

[0095] The non-high-quality data analysis data set analysis strategy determination module is used to determine the non-high-quality data set analysis strategy and analyze the non-high-quality data in the non-high-quality data analysis set.

[0096] The analysis strategy of the non-high-quality data set is: read the corresponding non-high-quality data from the storage location, perform anomaly detection on the non-high-quality data, analyze it if no anomaly is detected, and return to step S3 if anomaly is detected, and re-acquire the stored non-high-quality data after distributed storage. Specifically:

[0097] 1. Scan the non-high-quality data and determine the data points of the non-high-quality data.

[0098] The non-high-quality data X = {x1, x2, ...xi, ...xN} includes multiple data points, N is the size of the non-high-quality data, xi is a single data point in the non-high-quality data, and each data point contains multiple elements, denoted as xi1, xi2, ...xij, ...xiD, where each data point xi is a D-tuple of category A = (A1, A2, ...AD), A1, A2, ...AD represent different attributes of the elements, and each element can have the same or different attributes. Category A is only a general term and does not mean that it belongs to category A.

[0099] 2. Perform point division on the data points of non-high-quality data to determine whether there are abnormal data points. Specifically, point division involves dividing different data points into point sets. If there are data points that do not belong to the point set, they are considered abnormal data points. Specifically:

[0100] ①. Determine the initial center point and the initial point set.

[0101] A data point among multiple data points is randomly selected as an initial center point, and an initial point set is determined according to the initial center point.

[0102] The point set is in the r neighborhood, with the initial center point as the center. A circle with a radius of . Each data point represents a category. For example, data point xi is selected as the cluster center of category A.

[0103] ②. Update the initial point set and determine abnormal data points.

[0104] Continue to scan the data points other than the initial center point in the non-high-quality data. If the distance R(x A ,x i ) is less than or equal to If is the radius, the data point is put into the point set, completing a point set update, and the point set is continued to be updated by distance comparison.

[0105]

[0106] Among them, x A represents the initial center of category A, x i Represents a data point selected for distance calculation with the initial center, where x Aj Represents the initial center x of category A A Any element contained in x ij Represents data point x i Any element contained, n represents the number of elements contained in the data points in the selected category A, j is a natural number, (x Aj ,AD) represents the initial center x of category A A The attribute contained is the number of AD elements, (x ij ,AD) data point x of category A i The attribute contained is the number of AD elements.

[0107] Furthermore, if the distance between a data point and the initial center point is greater than If the radius is the radius, then the point needs to be used as another initial center point, the initial center point is defined as a new center point, and a new initial point set is determined based on the new center point. The determination method of the new initial point set refers to step S4221, which is not described in detail here. The distance between the remaining data points and the new center point is compared to update the new initial point set. The distance between the remaining data points and the new center point refers to the above formula, and the specific method is not described in detail here.

[0108] The update of one or more point sets is completed according to the above method, and each point set contains corresponding data points. If all the data points are included in the corresponding point set, it is considered that there are no abnormal data points, and non-quality data analysis is performed. The analysis method can refer to the regression analysis, correlation analysis and other methods in the prior art, and the specific method is not repeated here.

[0109] If there is a data point that is not included in one or more point sets, the data point is considered to be an abnormal data point. If an abnormal data point occurs, the stored non-high-quality data is re-read.

[0110] Since there may be multiple reads of non-quality data from multiple storage locations, it is also necessary to determine the access limit Q. If the access limit Q reaches the upper limit, the reading of non-quality data from one or more storage locations is stopped, and the non-quality data is stored in other storage locations.

[0111] The access quota Q is expressed as:

[0112]

[0113] where t p represents the number of times the non-quality data is accessed in the storage location p, P represents the total number of storage locations storing non-quality data, and v p represents the routing cost required to read non-quality data, ω P Represents the storage coefficient of the storage location p where non-quality data is stored. Generally speaking, the more times non-quality data is stored, the smaller the coefficient is, and vice versa. P The specific value is set in advance by the staff and is not limited here.

[0114] The present application also provides a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the artificial intelligence-based data analysis method.

[0115] The embodiments disclosed in the present invention provide a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed on a computer, the computer executes the above-mentioned data analysis method based on artificial intelligence.

[0116] An embodiment of the present invention provides a processor for processing the above-mentioned artificial intelligence-based data analysis method.

[0117] In the embodiment of the present invention, the processor may be an integrated circuit chip having the ability to process signals. The processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0118] The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor for execution, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and completes the steps of the above method in combination with its hardware.

[0119] The storage medium may be a memory, which may be, for example, a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memory.

[0120] Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct RAM bus random access memory (DRRAM).

[0121] This application has the following beneficial effects:

[0122] The present application can distinguish and divide data during the data analysis process, so as to perform different analysis and processing on different data, improve the accuracy of the analysis and speed up the execution time of the analysis.

[0123] Although the present application is described with reference to examples, this is for illustrative purposes only and is not intended to limit the present application, and changes, additions and / or deletions to the embodiments may be made without departing from the scope of the present application.

[0124] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A data analysis method based on artificial intelligence, characterized in that: The following steps are involved: Conduct data collection; Categorize the collected data and determine the data analysis set; Distributed storage of data analysis sets; After distributed storage, the data analysis strategy is determined, and the data analysis set is analyzed according to the data analysis strategy.

2. The artificial intelligence-based data analysis method according to claim 1, characterized in that: According to the classification of the collected data, determining the data analysis set includes the following sub-steps: Determine whether the collected data is missing; The collected data were categorized according to the missing results and the data analysis set was determined.

3. The artificial intelligence-based data analysis method according to claim 2, characterized in that: If the collected data is missing, the missing data will be used to form a non-high-quality data analysis set; if the collected data is not missing, the non-missing data will be used to form a high-quality data analysis set.

4. The artificial intelligence-based data analysis method according to claim 1, characterized in that: Determining whether data is missing includes performing data verification, data consistency checking, and data integrity constraint checking.

5. The artificial intelligence-based data analysis method according to claim 3, characterized in that: Distributed storage of data analysis sets includes interrupted storage of high-quality data analysis sets and continuous storage of low-quality data analysis sets.

6. A data analysis system based on artificial intelligence, characterized in that: Specifically include: a collection unit, an analysis set determination unit, a storage unit, and an analysis strategy determination unit; A collection unit, used for data collection; An analysis set determination unit, used to classify the collected data and determine a data analysis set; A storage unit for distributed storage of data analysis sets; The analysis strategy determination unit is used to determine the data analysis strategy and analyze the data analysis set according to the data analysis strategy.

7. The artificial intelligence-based data analysis system according to claim 6, characterized in that: The analysis set determination unit classifies the collected data, and determines the data analysis set including the following sub-steps: Determine whether the collected data is missing; The collected data were categorized according to the missing results and the data analysis set was determined.

8. The artificial intelligence-based data analysis system according to claim 7, characterized in that: If the collected data is missing, the analysis set determination unit forms a non-high-quality data analysis set with the missing data; if the collected data is not missing, the analysis set is formed with the non-missing data.

9. The artificial intelligence-based data analysis system according to claim 6, characterized in that: The analysis set determination unit determines whether the data is missing, including performing data verification, data consistency check and data integrity constraint check.

10. The artificial intelligence-based data analysis system according to claim 8, characterized in that: The storage unit performs distributed storage of the data analysis sets, including interrupting storage of high-quality data analysis sets and continuously storing low-quality data analysis sets.

Images