A multi-data storage docking method based on AI platform

Abstract

The present invention belongs to the technical field of data storage docking. The present invention provides a multi-data storage docking method based on an AI platform, including: acquiring multiple storage source data formats, and comparing the multiple storage source data formats with a target data format to obtain a matching compatibility value, and comparing it with the matching compatibility value; if the matching compatibility value is less than a matching compatibility threshold, generating a matching compatibility low signal, and marking the storage source data corresponding to the generated matching compatibility low signal as low-compatible data; analyzing and processing the low-compatible data to obtain a compatible conversion value, and comparing it with a compatible conversion threshold; if the compatible conversion value is greater than the compatible conversion threshold, generating a complex conversion signal, and marking the low-compatible data corresponding to the generated complex conversion signal as complex conversion data, quickly identifying the compatibility between different storage source data formats and the target data format, and warning of the difficulty of data conversion.

Description

Technical Field

[0001] The present invention belongs to the technical field of data storage docking, and specifically is a multi-data storage docking method based on an AI platform. Background Art

[0002] With the continuous development of artificial intelligence technology, AI platforms are being used more and more widely in all walks of life. These platforms need to process and analyze large amounts of data, so the demand for data storage is also growing. Traditional data storage methods often cannot meet the AI ​​platform's requirements for efficiency, scalability, and compatibility. Therefore, it is necessary to develop new multi-data storage docking methods.

[0003] In the prior art, different user storage systems have differences in data formats, communication protocols, interfaces, etc., which lead to the inability to adapt to different storage systems and require data conversion operations to achieve multi-data storage docking. However, multiple data in different user storage systems differ in different aspects. Therefore, by analyzing the similarities of multiple data in update time nodes and update frequencies, the synchronously converted data is extracted to solve the problem of data synchronous conversion of complex data during storage docking, and the synchronous conversion time of complex data converted within the synchronous conversion group during storage docking is determined.

[0004] To this end, the present invention provides a multi-data storage docking method based on an AI platform. Summary of the invention

[0005] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0006] The technical solution adopted by the present invention to solve its technical problem is:

[0007] A multi-data storage docking method based on an AI platform, comprising:

[0008] Step 1: Acquire multiple storage source data formats, and compare the multiple storage source data formats with the target data format to obtain content matching values ​​and position matching values, analyze and process the content matching values ​​and the position matching values, output a matching compatibility value, and compare it with the matching compatibility value. If the matching compatibility value is less than the matching compatibility threshold, generate a matching compatibility low signal, and mark the storage source data corresponding to the generated matching compatibility low signal as low compatible data;

[0009] Step 2: Based on the low-compatible data, the low-compatible data is analyzed and processed to obtain a compatible conversion value, and compared with the compatible conversion threshold. If the compatible conversion value is greater than the compatible conversion threshold, a conversion complex signal is generated, and the low-compatible data corresponding to the generated conversion complex signal is marked as conversion complex data;

[0010] Step 3: Based on the conversion complex signal, multiple conversion complex data are analyzed and processed to obtain periodic similarity groups, and the periodic similarity groups are analyzed and processed to obtain node comparison values ​​and frequency comparison values, the node comparison values ​​and the frequency comparison values ​​are processed, and the classification conversion values ​​are output to obtain classification conversion values, and the classification conversion values ​​are compared with the classification conversion threshold. If the classification conversion value is less than or equal to the classification conversion threshold, the periodic similarity group with higher similarity in data update and update frequency is marked as a synchronous conversion group;

[0011] Step 4: Based on the synchronous conversion group, obtain the conversion optimization coefficient and complete the optimization work of converting complex data within the synchronous conversion group.

[0012] As a further solution of the present invention, the content matching value is obtained by:

[0013] The storage source data format includes the source field position and the source field character, and the target data format includes the target field position and the target field character;

[0014] Compare the source field characters with the target field characters. The comparison process is as follows:

[0015] If the characters in the source field coincide with the characters in the target field, the field with the coincident characters will be marked as a field with coincident characters;

[0016] If the characters in the source field do not coincide with the characters in the target field, the field with inconsistent characters will be marked as a field with inconsistent characters;

[0017] Count the number of fields with overlapping characters, calculate the ratio of the number of fields with overlapping characters to the total number of target fields, and obtain the content matching value.

[0018] As a further solution of the present invention, a method for obtaining matching compatible values ​​is as follows:

[0019] Compare the source field position with the target field position. The comparison process is as follows:

[0020] Rank the source field positions to obtain the source field ranking;

[0021] Rank the target field positions to obtain the target field ranking;

[0022] Select the ranking of the field with overlapping characters in the source data in the source field ranking to obtain the source overlapping ranking;

[0023] Select the ranking of the field with overlapping characters in the target data within the ranking of the target field to obtain the target overlapping ranking;

[0024] Subtract the source coincidence ranking from the target coincidence ranking and take the absolute value to obtain the position matching value;

[0025] The content matching value is compared with the position matching value to obtain the matching compatibility value.

[0026] As a further solution of the present invention, the misaligned field is obtained in the following manner:

[0027] Get the number of fields with non-overlapping characters, calculate the ratio of the number of fields with non-overlapping characters to the total number of fields in the low-compatible data, and get the field difference value;

[0028] Extract the number of characters in the character non-overlapping field, calculate the ratio of the number of characters in the character non-overlapping field to the total number of characters in the low-compatible data, and obtain the character difference value;

[0029] The field difference value is multiplied by the character difference value to obtain the difference degree value;

[0030] The difference value is compared with the position difference threshold, and the process is as follows:

[0031] If the position difference value is greater than or equal to the position difference threshold, that is, the position difference is large, then the character overlap field corresponding to the large position difference is marked as a misplaced field.

[0032] As a further solution of the present invention, the compatible conversion value is obtained by:

[0033] Obtain the number of misplaced fields, calculate the ratio of the number of misplaced fields to the total number of fields in the low-compatible data, and obtain the field misplacement value;

[0034] Obtain the number of characters in the misplaced field, calculate the ratio of the number of characters in the misplaced field to the total number of characters in the low-compatible data, and obtain the character misplacement value;

[0035] Multiply the field misalignment value by the character misalignment value to calculate the misalignment degree value;

[0036] The difference degree value and the misalignment degree value are added together to obtain a compatible conversion value.

[0037] As a further solution of the present invention, the data processing time is obtained as follows:

[0038] Based on the conversion of complex data, the data collection time point and the data storage time point are obtained, and the data storage time point is subtracted from the data collection time point to obtain the data processing time.

[0039] As a further solution of the present invention, the method of obtaining the period similarity group is:

[0040] Integrate multiple complex data sets into one complex data set;

[0041] Extract any two transformed complex data from the complex data set as a complex data group;

[0042] In the complex data group, respectively obtain the data processing time corresponding to two converted complex data in the complex data group, and subtract the data processing time corresponding to the two converted complex data in the complex data group to obtain a processing deviation value;

[0043] The process deviation value is compared to the process deviation threshold as follows:

[0044] If the processing deviation value is less than the processing deviation threshold, that is, the processing cycles of the two converted complex data in the complex data group are similar, and the complex data groups with similar processing cycles are marked as cycle-similar groups.

[0045] As a further solution of the present invention, the classification conversion value is obtained by:

[0046] Get the data update time point in the period-similar group, subtract the data update time points corresponding to the two conversion data, take the absolute value, and get the update node difference;

[0047] Calculate the variance of all updated node differences to obtain the node comparison value;

[0048] Obtain the data update frequency within the group with similar cycles, subtract the data update frequencies corresponding to the two converted data, take the absolute value, and obtain the update frequency difference;

[0049] Calculate the ratio of the update frequency difference to the update frequency difference threshold to obtain a frequency comparison value;

[0050] The frequency comparison value and the node comparison value are added together to obtain the classification conversion value.

[0051] As a further solution of the present invention, the conversion optimization coefficient is obtained as follows:

[0052] Substitute the categorical conversion value into the formula: , calculate the conversion optimization coefficient K, where G represents the current classification conversion value, Expressed as the classification conversion threshold.

[0053] As a further solution of the present invention, the optimization process of converting complex data within a synchronous conversion group is as follows:

[0054] Substitute the conversion optimization factor K into the formula: , calculate the optimized synchronous conversion time ,in, It represents the current classification conversion value, and K represents the conversion optimization coefficient.

[0055] The beneficial effects of the present invention are as follows:

[0056] (1) The present invention obtains multiple storage source data formats, and compares the multiple storage source data formats with the target data format to obtain a matching compatibility value, and compares it with the matching compatibility value. If the matching compatibility value is less than the matching compatibility threshold, a matching compatibility low signal is generated, and the storage source data corresponding to the generated matching compatibility low signal is marked as low-compatible data. The low-compatible data is analyzed and processed to obtain a compatible conversion value, and compared with the compatible conversion threshold. If the compatible conversion value is greater than the compatible conversion threshold, a conversion complex signal is generated, and the low-compatible data corresponding to the generated conversion complex signal is marked as conversion complex data, so as to quickly identify the compatibility between different storage source data formats and the target data format, and warn of the difficulty of data conversion.

[0057] (2) Based on the conversion complex signal, the present invention analyzes and processes multiple conversion complex data to obtain periodic similarity groups, and analyzes and processes the periodic similarity groups to obtain node comparison values ​​and frequency comparison values. The node comparison values ​​and frequency comparison values ​​are processed and output to obtain classification conversion values. The classification conversion values ​​are compared with the classification conversion threshold. If the classification conversion value is less than or equal to the classification conversion threshold, the periodic similarity group with higher similarity in data update and update frequency is marked as a synchronous conversion group. Thus, by analyzing the similarity of multiple data in update time nodes and update frequencies, synchronously converted data is extracted, thereby solving the problem of data synchronous conversion of conversion complex data in the storage docking process;

[0058] (3) The present invention optimizes the conversion time of complex data converted within the synchronous conversion group, substitutes the classified conversion value into the formula, calculates the conversion optimization coefficient, and calculates the optimized synchronous conversion time through the formula based on the conversion optimization coefficient, thereby determining the synchronous conversion time of complex data converted within the synchronous conversion group when performing storage docking. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention will be further described below in conjunction with the accompanying drawings.

[0060] Figure 1 It is a flow chart of judging data compatibility and the difficulty of data conversion in the steps of a multi-data storage docking method based on an AI platform of the present invention;

[0061] Figure 2 This is a flowchart of the steps of determining synchronous data conversion and conversion optimization in a multi-data storage docking method based on an AI platform of the present invention. DETAILED DESCRIPTION

[0062] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0063] Example 1

[0064] like Figure 1-Figure 2 As shown, a multi-data storage docking method based on an AI platform according to an embodiment of the present invention includes:

[0065] Step 1: Acquire multiple storage source data formats, and compare the multiple storage source data formats with the target data format to obtain a matching compatibility value, and compare it with the matching compatibility value. If the matching compatibility value is less than the matching compatibility threshold, generate a matching compatibility low signal, and mark the storage source data corresponding to the generated matching compatibility low signal as low compatible data;

[0066] Specifically, the storage source data format includes the source field position and the source field character, and the target data format includes the target field position and the target field character;

[0067] Compare the source field characters with the target field characters. The comparison process is as follows:

[0068] If the characters in the source field coincide with the characters in the target field, the field with the coincident characters will be marked as a field with coincident characters;

[0069] If the characters in the source field do not coincide with the characters in the target field, the field with inconsistent characters will be marked as a field with inconsistent characters;

[0070] Count the number of fields with overlapping characters, calculate the ratio of the number of fields with overlapping characters to the total number of target fields, and obtain the content matching value;

[0071] Compare the source field position with the target field position. The comparison process is as follows:

[0072] Rank the source field positions to obtain the source field ranking;

[0073] Rank the target field positions to obtain the target field ranking;

[0074] Select the ranking of the field with overlapping characters in the source data in the source field ranking to obtain the source overlapping ranking;

[0075] Select the ranking of the field with overlapping characters in the target data within the ranking of the target field to obtain the target overlapping ranking;

[0076] Subtract the source coincidence ranking from the target coincidence ranking and take the absolute value to obtain the position matching value;

[0077] Calculate the ratio of the content matching value to the position matching value to obtain a matching compatibility value;

[0078] It can be understood that the meaning of the matching compatibility value is: through the evaluation of content similarity and position similarity, it reflects the matching compatibility between the stored source data and the target data in terms of format. Specifically, the larger the value, the stronger the matching and compatibility between the source field and the target field in terms of content and position. On the contrary, the smaller the value, the weaker the matching and compatibility between the source field and the target field in terms of content and position.

[0079] The matching compatibility value is compared with the matching compatibility threshold. The comparison process is as follows:

[0080] If the matching compatibility value is greater than or equal to the matching compatibility threshold, the matching compatibility between the source field and the target field in terms of content and position is relatively strong, a matching compatibility high signal is generated, and the stored source data corresponding to the generated matching compatibility high signal is marked as highly compatible data;

[0081] If the matching compatibility value is less than the matching compatibility threshold, the matching compatibility between the source field and the target field is weak in content and position, a matching compatibility low signal is generated, and the stored source data corresponding to the generated matching compatibility low signal is marked as low-compatible data;

[0082] Step 2: Based on the low-compatible data, the low-compatible data is analyzed and processed to obtain a compatible conversion value, and compared with the compatible conversion threshold. If the compatible conversion value is greater than the compatible conversion threshold, a conversion complex signal is generated, and the low-compatible data corresponding to the generated conversion complex signal is marked as conversion complex data;

[0083] In some embodiments, when low-compatibility data is obtained, a difference degree value and a misalignment degree value are obtained, and the difference degree value and the misalignment degree value are analyzed and processed to obtain a compatible conversion value. The specific process is as follows:

[0084] Get the number of fields with non-overlapping characters, calculate the ratio of the number of fields with non-overlapping characters to the total number of fields in the low-compatible data, and get the field difference value;

[0085] Extract the number of characters in the character non-overlapping field, calculate the ratio of the number of characters in the character non-overlapping field to the total number of characters in the low-compatible data, and obtain the character difference value;

[0086] The field difference value is multiplied by the character difference value to obtain the difference degree value;

[0087] The difference value is compared with the position difference threshold, and the process is as follows:

[0088] If the position difference value is greater than or equal to the position difference threshold, that is, the position difference is large, then the character overlap field corresponding to the large position difference is marked as a misplaced field;

[0089] If the position difference value is less than the position difference threshold, that is, the position difference is small, then the character overlap field corresponding to the small position difference is marked as a non-misplaced field;

[0090] Obtain the number of misplaced fields, calculate the ratio of the number of misplaced fields to the total number of fields in the low-compatible data, and obtain the field misplacement value;

[0091] Obtain the number of characters in the misplaced field, calculate the ratio of the number of characters in the misplaced field to the total number of characters in the low-compatible data, and obtain the character misplacement value;

[0092] Multiply the field misalignment value by the character misalignment value to calculate the misalignment degree value;

[0093] The difference degree value and the misalignment degree value are added and summed to obtain a compatible conversion value;

[0094] It can be understood that the meaning of the compatible conversion value is: a quantitative evaluation of the incompatibility between two sets of data (source data and target data), reflecting the differences between the source data and the target data in terms of fields and characters. Specifically, if this value is larger, it means that the differences between the source data and the target data in terms of fields and characters are larger, and if this value is smaller, it means that the differences between the source data and the target data in terms of fields and characters are smaller.

[0095] The compatible transition value is compared to the compatible transition threshold as follows:

[0096] If the compatible conversion value is greater than the compatible conversion threshold, it means that the source data and the target data are quite different in terms of fields and characters, and a conversion complex signal is generated, and the low compatible data corresponding to the generated conversion complex signal is marked as conversion complex data;

[0097] If the compatible conversion value is less than or equal to the compatible conversion threshold, it means that the differences between the source data and the target data in terms of fields and characters are small, and a signal that the conversion is not complex is generated;

[0098] The specific implementation scheme of the embodiment of the present invention is: obtain multiple storage source data formats, and compare the multiple storage source data formats with the target data format to obtain a matching compatibility value, and compare it with the matching compatibility value. If the matching compatibility value is less than the matching compatibility threshold, generate a matching compatibility low signal, and mark the storage source data corresponding to the generated matching compatibility low signal as low compatibility data, analyze and process the low compatibility data to obtain a compatible conversion value, and compare it with the compatible conversion threshold. If the compatible conversion value is greater than the compatible conversion threshold, generate a conversion complex signal, and mark the low compatibility data corresponding to the generated conversion complex signal as conversion complex data, quickly identify the compatibility between different storage source data formats and the target data format, and warn of the difficulty of data conversion.

[0099] Example 2

[0100] like Figure 1-Figure 2 As shown, based on Example 1, a multi-data storage docking method based on an AI platform according to an embodiment of the present invention includes:

[0101] Step 3: Based on the conversion complex signal, multiple conversion complex data are analyzed and processed to obtain periodic similarity groups, and the periodic similarity groups are analyzed and processed to obtain node comparison values ​​and frequency comparison values, the node comparison values ​​and the frequency comparison values ​​are processed, and the classification conversion values ​​are output to obtain classification conversion values, and the classification conversion values ​​are compared with the classification conversion threshold. If the classification conversion value is less than or equal to the classification conversion threshold, the periodic similarity group with higher similarity in data update and update frequency is marked as a synchronous conversion group;

[0102] Specifically, multiple transformed complex data are integrated into one complex data set;

[0103] Extract any two transformed complex data from the complex data set as a complex data group;

[0104] In the complex data group, respectively obtain the data processing time corresponding to two converted complex data in the complex data group, and subtract the data processing time corresponding to the two converted complex data in the complex data group to obtain a processing deviation value;

[0105] The data processing time is obtained as follows:

[0106] Based on the conversion of complex data, the data collection time point and the data storage time point are obtained, and the data storage time point is subtracted from the data collection time point to obtain the data processing time;

[0107] The process deviation value is compared to the process deviation threshold as follows:

[0108] If the processing deviation value is greater than the processing deviation threshold, it means that there is a difference between the two transformed complex data processing cycles in the complex data group;

[0109] If the processing deviation value is less than the processing deviation threshold, that is, the processing cycles of the two converted complex data in the complex data group are similar, and the complex data groups with similar processing cycles are marked as cycle-similar groups;

[0110] Get the data update time point in the period-similar group, subtract the data update time points corresponding to the two conversion data, take the absolute value, and get the update node difference;

[0111] Calculate the variance of all updated node differences to obtain the node comparison value;

[0112] Obtain the data update frequency within the group with similar cycles, subtract the data update frequencies corresponding to the two converted data, take the absolute value, and obtain the update frequency difference;

[0113] Calculate the ratio of the update frequency difference to the update frequency difference threshold to obtain a frequency comparison value;

[0114] The frequency comparison value and the node comparison value are added and summed to obtain a classification conversion value;

[0115] It can be understood that the meaning of the classification conversion value is: a comprehensive indicator for evaluating the consistency or similarity of two conversion complex data in a period-similar group in terms of data update and update frequency. Specifically, if this value is larger, it means that the two conversion complex data in a period-similar group have a lower degree of similarity in terms of data update and update frequency. If this value is smaller, it means that the two conversion complex data in a period-similar group have a higher degree of similarity in terms of data update and update frequency.

[0116] The collation conversion value is compared to the collation conversion threshold as follows:

[0117] If the classification conversion value is greater than the classification conversion threshold, it means that the two conversion complex data in the period similarity group have low similarity in data update and update frequency, and the period similarity group with low similarity in data update and update frequency is marked as an asynchronous conversion group;

[0118] If the classification conversion value is less than or equal to the classification conversion threshold, it means that the two complex conversion data in the periodic similarity group are highly similar in data update and update frequency, and the periodic similarity group with high similarity in data update and update frequency is marked as a synchronous conversion group.

[0119] The specific implementation scheme of the embodiment of the present invention is: based on the conversion complex signal, multiple conversion complex data are analyzed and processed to obtain periodic similarity groups, and the periodic similarity groups are analyzed and processed to obtain node comparison values ​​and frequency comparison values, the node comparison values ​​and the frequency comparison values ​​are processed, and the classification conversion values ​​are output to obtain classification conversion values, and the classification conversion values ​​are compared with the classification conversion threshold. If the classification conversion value is less than or equal to the classification conversion threshold, the periodic similarity group with higher similarity in data update and update frequency is marked as a synchronous conversion group, so as to extract synchronously converted data by analyzing the similarity of multiple data in update time nodes and update frequencies, thereby solving the problem of data synchronization conversion of conversion complex data during storage docking.

[0120] Example 3

[0121] like Figure 1-Figure 2 As shown, based on Example 1 and Example 2, a multi-data storage docking method based on an AI platform described in an embodiment of the present invention includes:

[0122] Step 4: Based on the synchronous conversion group, obtain the conversion optimization coefficient and complete the optimization of complex data conversion in the synchronous conversion group;

[0123] In some embodiments, the classification conversion value is substituted into the formula: , calculate the conversion optimization coefficient K, where G represents the current classification conversion value, It is expressed as the classification conversion threshold;

[0124] Substitute the conversion optimization factor K into the formula: , calculate the optimized synchronous conversion time ,in, It represents the current classification conversion value, and K represents the conversion optimization coefficient.

[0125] The specific implementation scheme of the embodiment of the present invention is: optimizing the conversion time of complex data converted within the synchronous conversion group, substituting the classified conversion value into the formula, calculating the conversion optimization coefficient, and calculating the optimized synchronous conversion time through the formula based on the conversion optimization coefficient, thereby determining the synchronous conversion time of complex data converted within the synchronous conversion group during storage docking.

[0126] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A multi-data storage docking method based on an AI platform, characterized in that: include: Step 1: Acquire multiple storage source data formats, and compare the multiple storage source data formats with the target data format to obtain content matching values ​​and position matching values, analyze and process the content matching values ​​and the position matching values, output a matching compatibility value, and compare it with the matching compatibility value. If the matching compatibility value is less than the matching compatibility threshold, generate a matching compatibility low signal, and mark the storage source data corresponding to the generated matching compatibility low signal as low compatible data; The content matching value is obtained as follows: The storage source data format includes the source field position and the source field character, and the target data format includes the target field position and the target field character; Compare the source field characters with the target field characters. The comparison process is as follows: If the characters in the source field coincide with the characters in the target field, the field with the coincident characters will be marked as a field with coincident characters; If the characters in the source field do not coincide with the characters in the target field, the field with inconsistent characters will be marked as a field with inconsistent characters; Count the number of fields with overlapping characters, calculate the ratio of the number of fields with overlapping characters to the total number of target fields, and obtain content matching; Step 2: Based on the low-compatible data, the low-compatible data is analyzed and processed to obtain a compatible conversion value, and compared with the compatible conversion threshold. If the compatible conversion value is greater than the compatible conversion threshold, a conversion complex signal is generated, and the low-compatible data corresponding to the generated conversion complex signal is marked as conversion complex data; Step 3: Based on the conversion complex signal, multiple conversion complex data are analyzed and processed to obtain periodic similarity groups, and the periodic similarity groups are analyzed and processed to obtain node comparison values ​​and frequency comparison values, the node comparison values ​​and the frequency comparison values ​​are processed, and the classification conversion values ​​are output to obtain classification conversion values, and the classification conversion values ​​are compared with the classification conversion threshold. If the classification conversion value is less than or equal to the classification conversion threshold, the periodic similarity group with higher similarity in data update and update frequency is marked as a synchronous conversion group; Step 4: Based on the synchronous conversion group, obtain the conversion optimization coefficient and complete the optimization of complex data conversion in the synchronous conversion group; The conversion optimization coefficient is obtained as follows: Substitute the categorical conversion value into the formula: , calculate the conversion optimization coefficient K, where G represents the current classification conversion value, It is expressed as the classification conversion threshold; The optimization process of converting complex data within a synchronous conversion group is as follows: Substitute the conversion optimization factor K into the formula: , calculate the optimized synchronous conversion time ,in, It represents the current classification conversion value, and K represents the conversion optimization coefficient.

2. According to claim 1, a multi-data storage docking method based on an AI platform is characterized in that: How to obtain matching compatible values: Compare the source field position with the target field position. The comparison process is as follows: Rank the source field positions to obtain the source field ranking; Rank the target field positions to obtain the target field ranking; Select the ranking of the field with overlapping characters in the source data in the source field ranking to obtain the source overlapping ranking; Select the ranking of the field with overlapping characters in the target data within the ranking of the target field to obtain the target overlapping ranking; Subtract the source coincidence ranking from the target coincidence ranking and take the absolute value to obtain the position matching value; The content matching value is compared with the position matching value to obtain the matching compatibility value.

3. The multi-data storage docking method based on AI platform according to claim 2 is characterized in that: The way to obtain the misplaced field is: Get the number of fields with non-overlapping characters, calculate the ratio of the number of fields with non-overlapping characters to the total number of fields in the low-compatible data, and get the field difference value; Extract the number of characters in the character non-overlapping field, calculate the ratio of the number of characters in the character non-overlapping field to the total number of characters in the low-compatible data, and obtain the character difference value; The field difference value is multiplied by the character difference value to obtain the difference degree value; The difference value is compared with the position difference threshold, and the process is as follows: If the position difference value is greater than or equal to the position difference threshold, that is, the position difference is large, then the character overlap field corresponding to the large position difference is marked as a misplaced field.

4. The multi-data storage docking method based on an AI platform according to claim 3, characterized in that: The compatible conversion value is obtained as follows: Obtain the number of misplaced fields, calculate the ratio of the number of misplaced fields to the total number of fields in the low-compatible data, and obtain the field misplacement value; Obtain the number of characters in the misplaced field, calculate the ratio of the number of characters in the misplaced field to the total number of characters in the low-compatible data, and obtain the character misplacement value; Multiply the field misalignment value by the character misalignment value to calculate the misalignment degree value; The difference degree value and the misalignment degree value are added together to obtain a compatible conversion value.

5. The multi-data storage docking method based on an AI platform according to claim 1, characterized in that: The data processing time is obtained as follows: Based on the conversion of complex data, the data collection time point and the data storage time point are obtained, and the data storage time point is subtracted from the data collection time point to obtain the data processing time.

6. The multi-data storage docking method based on AI platform according to claim 5, characterized in that: The method for obtaining a period similarity group is as follows: Integrate multiple complex data sets into one complex data set; Extract any two transformed complex data from the complex data set as a complex data group; In the complex data group, respectively obtain the data processing time corresponding to two converted complex data in the complex data group, and subtract the data processing time corresponding to the two converted complex data in the complex data group to obtain a processing deviation value; The process deviation value is compared to the process deviation threshold as follows: If the processing deviation value is less than the processing deviation threshold, that is, the processing cycles of the two converted complex data in the complex data group are similar, and the complex data groups with similar processing cycles are marked as cycle-similar groups.

7. The multi-data storage docking method based on AI platform according to claim 6 is characterized in that: The way to obtain the classification conversion value is: Get the data update time point in the period-similar group, subtract the data update time points corresponding to the two conversion data, take the absolute value, and get the update node difference; Calculate the variance of all updated node differences to obtain the node comparison value; Obtain the data update frequency within the group with similar cycles, subtract the data update frequencies corresponding to the two converted data, take the absolute value, and obtain the update frequency difference; Calculate the ratio of the update frequency difference to the update frequency difference threshold to obtain a frequency comparison value; The frequency comparison value and the node comparison value are added together to obtain the classification conversion value.

Images