A metadata scoring method, storage medium and system of a business object

Abstract

The application provides a metadata scoring method, a storage medium and a system of a business object, which comprises the following steps: obtaining table-level metadata and field-level metadata used for describing the business object; scoring each field-level metadata to obtain field scores of the field-level metadata; performing mean value calculation on the field scores of all field-level metadata contained in each table-level metadata to obtain field comprehensive scores of all field-level metadata contained in the table-level metadata; scoring each table-level metadata, wherein, in the process, the field comprehensive scores of all field-level metadata contained in the table-level metadata are used as scores of one table two-level scoring index of the table-level metadata, table scores of the table-level metadata are calculated, and mean value calculation is performed on the table scores of all table-level metadata to obtain a metadata comprehensive score of the business object.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a metadata scoring method of a business object, a storage medium and a system. BACKGROUND

[0002] In a metadata management system, a business object is usually described by multiple table-level metadata, and each table-level metadata usually contains multiple field-level metadata. These table-level metadata and field-level metadata are mainly applied to data retrieval and knowledge graph processing of the business object, and the data retrieval and knowledge graph processing results of the business object are affected by the quality of the table-level metadata and field-level metadata. In order to ensure the reliability of the data retrieval and knowledge graph processing results of the business object, it is necessary to ensure the reliability of the quality of the table-level metadata and field-level metadata used to describe the business object. Therefore, it is necessary to evaluate the quality of the table-level metadata and field-level metadata first. However, since the field-level metadata is contained in the table-level metadata, the quality of the field-level metadata will affect the quality of the table-level metadata. The existing metadata quality evaluation methods usually evaluate the quality of the table-level metadata and field-level metadata respectively, and do not consider the quality influence between the table-level metadata and field-level metadata, which makes it difficult to ensure the reliability of the data retrieval and knowledge graph processing results of the business object. SUMMARY

[0003] The technical problem to be solved by the present application is how to improve the reliability of data retrieval and knowledge graph processing of a business object.

[0004] To solve the above technical problem, the present application provides a metadata scoring method of a business object, comprising the following steps:

[0005] A. obtaining at least one table-level metadata used to describe the business object and at least one field-level metadata contained in each table-level metadata;

[0006] B. scoring each field-level metadata, specifically comprising the following steps B1-B4:

[0007] B1. obtaining at least two field-level first scoring indicators used to score the field-level metadata and at least one field-level second scoring indicator contained in each field-level first scoring indicator;

[0008] B2. obtaining the weight respectively given by a user to each field-level first scoring indicator and each field-level second scoring indicator;

[0009] B3. obtaining field original indicator data of each field-level second scoring indicator, and calculating the score of each field-level second scoring indicator by pre-processing each field original indicator data;

[0010] —B4. The field score of the current field level metadata is obtained by weighting the scores of the field secondary evaluation indexes according to the weights of the field secondary evaluation indexes and the weights of the field primary evaluation indexes;

[0011] C. The field comprehensive scores of all the field level metadata contained in the current table level metadata are obtained by averaging the field scores of the field level metadata, respectively;

[0012] D. Each table level metadata is scored, specifically including the following steps D1-D4:

[0013] —D1. At least two table primary evaluation indexes and at least one table secondary evaluation index contained in each table primary evaluation index are obtained for scoring the current table level metadata, and the field comprehensive score of all the field level metadata contained in the current table level metadata is taken as the score of one of the table secondary evaluation indexes;

[0014] —D2. The weights respectively given by the user to each table primary evaluation index and each table secondary evaluation index are obtained;

[0015] —D3. Table original index data of each table secondary evaluation index except the field comprehensive score is obtained, and the scores of the other table secondary evaluation indexes are obtained by pre-processing the table original index data, respectively;

[0016] —D4. The table score of the current table level metadata is obtained by weighting the scores of the table secondary evaluation indexes according to the weights of the table secondary evaluation indexes and the weights of the table primary evaluation indexes;

[0017] E. The table scores of all the table level metadata for describing the business object are averaged to obtain the metadata comprehensive score of the business object.

[0018] Preferably, in the step B1, the field primary evaluation indexes include a field level integrity index, and the field secondary evaluation indexes contained in the field level integrity index include a field level technical metadata integrity index, a field level management metadata integrity index and a field level business metadata integrity index.

[0019] Preferably, in the step B2, the sum of the weights of all the field primary evaluation indexes of the same field level metadata is 100%, and the sum of the weights of all the field secondary evaluation indexes contained in the same field primary evaluation index is 100%.

[0020] Preferably, in the step B4, the scores of the field secondary scoring indicators are weighted according to the weights of the field secondary scoring indicators to obtain the scores of the field primary scoring indicators of the field level metadata, and the scores of the field primary scoring indicators are weighted according to the weights of the field primary scoring indicators to obtain the field score of the field level metadata.

[0021] Preferably, in the step D1, the table primary scoring indicators include a table level integrity indicator and a field indicator, the table secondary scoring indicators contained in the table level integrity indicator include a table level technical metadata integrity indicator, a table level management metadata integrity indicator and a table level business metadata integrity indicator, and the field comprehensive score is taken as the table secondary scoring indicators contained in the field indicator.

[0022] Preferably, in the step D2, the sum of the weights of all the table primary scoring indicators of the same table level metadata is 100%, and the sum of the weights of all the table secondary scoring indicators contained in the same table primary scoring indicator is 100%.

[0023] Preferably, in the step D4, the scores of the table secondary scoring indicators are weighted according to the weights of the table secondary scoring indicators to obtain the scores of the table primary scoring indicators of the table level metadata, and the scores of the table primary scoring indicators are weighted according to the weights of the table primary scoring indicators to obtain the table score of the table level metadata.

[0024] Preferably, in the step E, after obtaining the metadata comprehensive score of the business object, the business object is rated according to the metadata comprehensive score of the business object.

[0025] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps in the metadata scoring method of the business object.

[0026] The application further provides a metadata scoring system of a business object, which comprises a computer readable storage medium and a processor connected to each other, and the computer readable storage medium is as described above.

[0027] The present application has the following beneficial effects: the present application first scores each field-level metadata to obtain field scores of each field-level metadata, then respectively calculates the mean of the field scores of all field-level metadata contained in each table-level metadata to obtain the field comprehensive scores of all field-level metadata contained in each table-level metadata, then scores each table-level metadata, in which process, the field comprehensive scores of all field-level metadata contained in the table-level metadata are taken as one of the scores of table two-level score indicators, then the scores of each table two-level score indicator are weighted calculated according to the corresponding indicator weight to obtain the table scores of the table-level metadata, then the mean of the table scores of all table-level metadata is calculated to obtain the metadata comprehensive score of the business object, so that the finally obtained metadata comprehensive score of the business object combines the table scores of all table-level metadata and the field scores of all field-level metadata, and comprehensively considers the quality influence between the table-level metadata and the field-level metadata for describing the business object, so that the reliability of data retrieval and knowledge graph processing using the business object with high comprehensive score can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a flowchart of a metadata scoring method of a business object;

[0029] Figure 2 is a diagram of a field one-level score indicator and a field two-level score indicator of field-level metadata;

[0030] Figure 3 is a weight distribution diagram of a field one-level score indicator and a field two-level score indicator of field-level metadata;

[0031] Figure 4 is a diagram of a table one-level score indicator and a table two-level score indicator of table-level metadata;

[0032] Figure 5 is a weight distribution diagram of a table one-level score indicator and a table two-level score indicator of table-level metadata;

[0033] Figure 6 is a diagram of the corresponding relationship between the metadata comprehensive score and the rating of a business object. DETAILED DESCRIPTION

[0034] The present application will be further described in detail below in combination with specific embodiments.

[0035] The present embodiment provides a metadata scoring system of a business object, which comprises a computer readable storage medium and a processor connected to each other, the computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the metadata scoring method of the business object as Figure 1The metadata scoring method of the business object shown specifically comprises the following steps A, B, C, D and E.

[0036] A. Obtain at least one table-level metadata for describing the business object and at least one field-level metadata contained in each table-level metadata.

[0037] In the metadata management system, the business object is usually described by at least two table-level metadata, and each table-level metadata usually contains at least one field-level metadata. These table-level metadata and field-level metadata are mainly used for data retrieval and knowledge graph processing of the business object, and the data retrieval and knowledge graph processing results of the business object will be affected by the quality of the table-level metadata and field-level metadata. In the embodiment, in order to improve the reliability of the data retrieval and knowledge graph processing results of the business object, the business object is comprehensively scored in combination with all table-level metadata and all field-level metadata for describing the business object. Based on this, the scoring system obtains at least one table-level metadata for describing the business object and at least one field-level metadata contained in each table-level metadata. For example, the table-level metadata for describing a certain business object has two, which are the first table-level metadata and the second table-level metadata. The first table-level metadata contains two field-level metadata, which are the first field-level metadata and the second field-level metadata. The second table-level metadata contains two field-level metadata, which are the third field-level metadata and the fourth field-level metadata. In this way, the scoring system obtains the first table-level metadata and the second table-level metadata for describing the business object, and obtains the first field-level metadata and the second field-level metadata contained in the first table-level metadata, and obtains the third field-level metadata and the fourth field-level metadata contained in the second table-level metadata.

[0038] It should be noted that the number of table-level metadata for describing different business objects may be different, and the number of field-level metadata contained in each table-level metadata may also be different, which is determined according to the actual situation.

[0039] B. Score each field-level metadata.

[0040] The scoring system scores the first field-level metadata, the second field-level metadata, the third field-level metadata and the fourth field-level metadata obtained by it respectively. The scoring of each field-level metadata specifically comprises the following steps B1, B2, B3 and B4.

[0041] B1. Obtain at least two field-level first scoring indicators for scoring the field-level metadata and at least one field-level second scoring indicator contained in each field-level first scoring indicator.

[0042] In this embodiment, multiple scoring indicators at different levels are used to score field-level metadata. These multiple scoring indicators refer to at least two primary field scoring indicators and at least one secondary field scoring indicator included in each primary field scoring indicator. In this embodiment, if all field-level metadata has the same primary field scoring indicator and the same secondary field scoring indicator, then the scoring method for each field-level metadata is the same.

[0043] Taking the first field-level metadata as an example, the scoring of this field-level metadata uses five primary field scoring indicators: field-level completeness indicators, data standard indicators, quality rule indicators, quality issue indicators, and governance case indicators. Specifically: the field-level completeness indicators include three secondary field scoring indicators: field-level technical metadata completeness indicators, field-level management metadata completeness indicators, and field-level business metadata completeness indicators; the data standard indicators include one secondary field scoring indicator, specifically the standard compliance indicator; the quality rule indicators include one secondary field scoring indicator, specifically the association rule indicator; the quality issue indicators include one secondary field scoring indicator, specifically the quality issue percentage indicator; and the governance case indicators include one secondary field scoring indicator, specifically the case volume indicator. The scoring system obtains the five primary field scoring indicators used to score the first field-level metadata and the various secondary field scoring indicators contained in each primary field scoring indicator, constructing a scoring system as follows: Figure 2 The field-level metadata scoring index system is shown.

[0044] B2. Obtain the weights assigned by the user to the primary rating indicators and secondary rating indicators of each field.

[0045] After constructing the field-level metadata scoring indicator system, users can input the weights of each primary and secondary scoring indicator of the first field-level metadata item into the scoring system based on expert evaluations. The scoring system then obtains the weights assigned by the user to each primary and secondary scoring indicator of the first field-level metadata item, constructing a scoring system as follows: Figure 3The field-level evaluation model based on the balanced score card is shown. Among the field-level score indicators, the weight of the field-level integrity indicator is 60%, the weight of the data standard indicator is 15%, the weight of the quality rule indicator is 10%, the weight of the quality problem indicator is 10%, and the weight of the governance case indicator is 5%. Among the three field-level secondary score indicators included in the field-level integrity indicator, the weight of the field-level technical metadata integrity indicator is 80%, the weight of the field-level management metadata integrity indicator is 10%, and the weight of the field-level business metadata integrity indicator is 10%. The weight of the standard satisfaction indicator included in the data standard indicator is 100%. The weight of the correlation rule indicator included in the quality rule indicator is 100%. The weight of the quality problem proportion indicator included in the quality problem indicator is 100%. The weight of the case level indicator included in the governance case indicator is 100%.

[0046] It should be noted that the weights of the respective field-level score indicators and the respective field-level secondary score indicators given by the experts after evaluation may be different for different field-level metadata, but the sum of the weights of all field-level score indicators of the same field-level metadata is 100%, and the sum of the weights of all field-level secondary score indicators included in the same field-level score indicator is 100%.

[0047] B3. Obtain field original indicator data of each field-level secondary score indicator, and calculate scores of each field-level secondary score indicator by respectively pre-processing the field original indicator data.

[0048] After obtaining the weights respectively given by the user to the field-level score indicators and the field-level secondary score indicators, the scoring system obtains field original indicator data of each field-level secondary score indicator of the first field-level metadata, and then respectively adopts different preprocessing methods to calculate scores of each field-level secondary score indicator by using the obtained field original indicator data. Specifically:

[0049] For the field-level technical metadata integrity indicator, the field original indicator data obtained by the scoring system is the total number of field-level technical metadata and the number of non-empty field-level technical metadata in the first field-level metadata. The preprocessing method adopted is to calculate the ratio of the number of non-empty field-level technical metadata to the total number of field-level technical metadata. The calculation result is the score of the field-level technical metadata integrity indicator, i.e., the score of the field-level technical metadata integrity indicator = the number of non-empty field-level technical metadata / the total number of field-level technical metadata. Therefore, the score of the field-level technical metadata integrity indicator ranges from 0 to 1.

[0050] For the field-level management metadata integrity indicator, the field original indicator data obtained by the scoring system is the total number of field-level management metadata in the first item of field-level metadata and the number of non-empty field-level management metadata therein. The preprocessing method adopted is to calculate the ratio of the number of non-empty field-level management metadata to the total number of field-level management metadata. The calculation result is the score of the field-level management metadata integrity indicator, i.e., the score of the field-level management metadata integrity indicator = the number of non-empty field-level management metadata / the total number of field-level management metadata. The score range of the field-level management metadata integrity indicator is [0, 1];

[0051] For the field-level business metadata integrity indicator, the field original indicator data obtained by the scoring system is the total number of field-level business metadata in the first item of field-level metadata and the number of non-empty field-level business metadata therein. The preprocessing method adopted is to calculate the ratio of the number of non-empty field-level business metadata to the total number of field-level business metadata. The calculation result is the score of the field-level business metadata integrity indicator, i.e., the score of the field-level business metadata integrity indicator = the number of non-empty field-level business metadata / the total number of field-level business metadata. The score range of the field-level business metadata integrity indicator is [0, 1];

[0052] For the standard satisfaction indicator, the field original indicator data obtained by the scoring system is data reflecting whether the field in the first item of field-level metadata has mapped a data standard. The preprocessing method adopted is to determine whether the field has mapped a data standard according to the data. If yes, the score of the standard satisfaction indicator is 1. If no, the score of the standard satisfaction indicator is 0.

[0053] For the association rule indicator, the field original indicator data obtained by the scoring system is data reflecting whether the field in the first item of field-level metadata has associated a data quality audit rule. The preprocessing method adopted is to determine whether the field has associated a data quality audit rule according to the data. If yes, the score of the association rule indicator is 1. If no, the score of the association rule indicator is 0.

[0054] For the quality problem proportion indicator, the field original indicator data obtained by the scoring system is the total number of fields in the first item of field-level metadata and the number of problem fields therein. The preprocessing method adopted is to calculate the ratio of the number of fields without problems to the total number of fields. The calculation result is the score of the quality problem proportion indicator, i.e., the score of the quality problem proportion indicator = (the total number of fields - the number of problem fields) / the total number of fields. The score range of the quality problem proportion indicator is [0, 1];

[0055] For the case magnitude index, the field original index data obtained by the scoring system is the number of cases associated with the field in the first field-level metadata. The preprocessing method adopted is to set the score of the case magnitude index according to the number of cases associated with the field. Specifically, if the field is associated with more than one case, the score of the case magnitude index is 1; if the field is associated with one case, the score of the case magnitude index is 0.6; and if the field is not associated with any case, the score of the case magnitude index is 0.

[0056] B4. The scores of the field secondary scoring indicators are weighted and calculated according to the weights of the field secondary scoring indicators and the weights of the field primary scoring indicators to obtain the field score of the field-level metadata.

[0057] After obtaining the scores of the field secondary scoring indicators of the first field-level metadata, the scoring system weights and calculates the scores of the field secondary scoring indicators according to the weights of the field secondary scoring indicators and the weights of the field primary scoring indicators in the field secondary scoring indicators of the first field-level metadata to obtain the field score of the first field-level metadata. Specifically: Figure 3

[0058] In this embodiment, among the field secondary scoring indicators of the first field-level metadata: the number of non-empty field-level technical metadata / total number of field-level technical metadata = 0.8, so the score of the field-level technical metadata integrity indicator is 0.8; the number of non-empty field-level management metadata / total number of field-level management metadata = 0.7, so the score of the field-level management metadata integrity indicator is 0.7; the number of non-empty field-level business metadata / total number of field-level business metadata = 0.9, so the score of the field-level business metadata integrity indicator is 0.9; the field has mapped a data standard, so the score of the standard satisfaction indicator is 1; the field is not associated with a data quality audit rule, so the score of the associated rule indicator is 0; (total number of fields-number of problem fields) / total number of fields = 0.8, so the score of the quality problem proportion indicator is 0.8; the field is associated with one case, so the score of the case magnitude indicator is 0.6.

[0059] Therefore, the scoring system first weights and calculates the scores of the field secondary scoring indicators according to the weights of the field secondary scoring indicators to obtain the scores of the field primary scoring indicators. The calculation formula is:

[0060] Field primary scoring indicator score = ∑field secondary scoring indicator score*weight of field secondary scoring indicator;

[0061] ​The score of the field-level integrity indicator among the field-level first-level score indicators of the first item of field-level metadata is: the score of the field-level technical metadata integrity indicator * the weight of the field-level technical metadata integrity indicator + the score of the field-level management metadata integrity indicator * the weight of the field-level management metadata integrity indicator + the score of the field-level business metadata integrity indicator * the weight of the field-level business metadata integrity indicator = 0.8 * 80% + 0.7 * 10% + 0.9 * 10% = 0.8; the score of the data standard indicator is: the score of the standard satisfaction indicator * the weight of the standard satisfaction indicator = 1 * 100% = 1; the score of the quality rule indicator is: the score of the associated rule indicator * the weight of the associated rule indicator = 0 * 100% = 0; the score of the quality problem indicator is: the score of the quality problem proportion indicator * the weight of the quality problem proportion indicator = 0.8 * 100% = 0.8; and the score of the governance case indicator is: the score of the case quantity level indicator * the weight of the case quantity level indicator = 0.6 * 100% = 0.6.

[0062] Then, the scoring system calculates the field score of the first item of field-level metadata by weighting the scores of the field-level first-level score indicators according to the weights of the field-level first-level score indicators, and the calculation formula is:

[0063] The field score of the field-level metadata = ∑ the score of the field-level first-level score indicator * the weight of the field-level first-level score indicator.

[0064] The field score of the first item of field-level metadata = the score of the field-level integrity indicator * the weight of the field-level integrity indicator + the score of the data standard indicator * the weight of the data standard indicator + the score of the quality rule indicator * the weight of the quality rule indicator + the score of the quality problem indicator * the weight of the quality problem indicator + the score of the governance case indicator * the weight of the governance case indicator = 0.8 * 60% + 1 * 15% + 0 * 10% + 0.8 * 10% + 0.6 * 5% = 0.74.

[0065] Similarly, since the field-level first-level score indicators of each item of field-level metadata in the embodiment are the same, and the field-level second-level score indicators of each item of field-level metadata are the same, the scoring manner for each item of field-level metadata is the same, the same scoring manner as the first item of field-level metadata can be used to score the second item of field-level metadata, the third item of field-level metadata and the fourth item of field-level metadata respectively, so that the field score of the second item of field-level metadata is for example 0.84, the field score of the third item of field-level metadata is for example 0.59, and the field score of the fourth item of field-level metadata is for example 0.76.

[0066] C. The field scores of all the field-level metadata contained in each item of table-level metadata are respectively calculated in mean, so as to obtain the field comprehensive score of all the field-level metadata contained in each item of table-level metadata.

[0067] In this embodiment, since the first item table level metadata contains the first item field level metadata and the second item field level metadata, the field score of the first item field level metadata and the field score of the second item field level metadata can be averaged to obtain the field comprehensive score of all the field level metadata contained in the first item table level metadata, and the field comprehensive score of all the field level metadata contained in the first item table level metadata = (the field score of the first item field level metadata + the field score of the second item field level metadata) / 2 = (0.74 + 0.84) / 2 = 0.79.

[0068] Since the second item table level metadata contains the third item field level metadata and the fourth item field level metadata, the field score of the third item field level metadata and the field score of the fourth item field level metadata can be averaged to obtain the field comprehensive score of all the field level metadata contained in the second item table level metadata, and the field comprehensive score of all the field level metadata contained in the second item table level metadata = (the field score of the third item field level metadata + the field score of the fourth item field level metadata) / 2 = (0.59 + 0.76) / 2 = 0.675.

[0069] D. Score each item table level metadata.

[0070] After scoring each item field level metadata respectively, the scoring system scores the first item table level metadata and the second item table level metadata obtained by the scoring system respectively, and scoring each item table level metadata specifically includes the following steps D1, D2, D3 and D4.

[0071] D1. Obtain at least two table level scoring indicators for scoring the table level metadata and at least one table level scoring indicator contained in each table level scoring indicator, and take the field comprehensive score of all the field level metadata contained in the table level metadata as one of the table level scoring indicators.

[0072] In this embodiment, different levels of multiple scoring indicators are used to score the table level metadata, wherein the different levels of multiple scoring indicators refer to at least two table level scoring indicators and at least one table level scoring indicator contained in each table level scoring indicator. In this embodiment, the table level scoring indicators of each item table level metadata are the same, and the table level scoring indicators of each item table level metadata are the same, so the scoring method of each item table level metadata is the same.

[0073] Taking the first table-level metadata item as an example, the scoring of this item uses two table-level scoring indicators: a table-level integrity indicator and a field indicator. The table-level integrity indicator includes three table-level scoring indicators: a table-level technical metadata integrity indicator, a table-level management metadata integrity indicator, and a table-level business metadata integrity indicator. The field indicator includes one table-level scoring indicator, specifically a comprehensive field scoring indicator, which is the comprehensive score of all field-level metadata included in the first table-level metadata item. The scoring system obtains the two table-level scoring indicators used to score the first table-level metadata item and the various table-level scoring indicators included in each table-level scoring indicator, constructing a scoring system as follows: Figure 4 The table-level metadata scoring index system is shown.

[0074] D2. Obtain the weights assigned by users to the primary and secondary rating indicators of each table.

[0075] After constructing the table-level metadata scoring index system, users can input the weights of each primary and secondary scoring index of the first table-level metadata item into the scoring system based on expert evaluations. The scoring system then obtains the weights assigned by the user to each primary and secondary scoring index of the first table-level metadata item, constructing a scoring index system as follows: Figure 5 The table-level evaluation model shown is based on the Balanced Scorecard. In the table-level scoring indicators, the table-level integrity indicator has a weight of 60%, and the field indicator has a weight of 40%. Among the three table-level scoring indicators included in the table-level integrity indicator, the table-level technical metadata integrity indicator has a weight of 80%, the table-level management metadata integrity indicator has a weight of 10%, and the table-level business metadata integrity indicator has a weight of 10%. The field indicator, which includes the comprehensive field scoring indicator, has a weight of 100%.

[0076] It should be noted that the weights of the various table-level scoring indicators and the various table-level scoring indicators given by experts after evaluation may differ for different table-level metadata. However, it is necessary to ensure that the sum of the weights of all table-level scoring indicators for the same table-level metadata is 100%, and the sum of the weights of all table-level scoring indicators included in the same table-level scoring indicator is 100%.

[0077] D3. Obtain the original index data of the secondary scoring indicators of each table other than the comprehensive score of the field, and preprocess the original index data of each table to calculate the scores of the secondary scoring indicators of the other tables.

[0078] After obtaining the weight respectively given by the user to the table first-level score indicator and the table second-level score indicator, since one of the table second-level score indicators (field comprehensive score indicator) is the field comprehensive score of all the field-level metadata contained in the first table-level metadata, the field comprehensive score of all the field-level metadata contained in the first table-level metadata is the score of the field comprehensive score indicator, i.e., the score of the field comprehensive score indicator of the first table-level metadata is 0.79. In addition, the scoring system obtains the table original indicator data of each table second-level score indicator except the field comprehensive score, i.e., obtains the table original indicator data of the table-level technical metadata integrity indicator, the table-level management metadata integrity indicator and the table-level business metadata integrity indicator, respectively, and then respectively pre-processes and calculates each table original indicator data to obtain the score of each table second-level score indicator. Specifically:

[0079] For the table-level technical metadata integrity indicator, the table original indicator data obtained by the scoring system is the total number of table-level technical metadata in the first table-level metadata and the number of non-empty table-level technical metadata therein, and the pre-processing manner adopted is to calculate the ratio of the number of non-empty table-level technical metadata to the total number of table-level technical metadata, and the calculation result is the score of the table-level technical metadata integrity indicator, i.e., the score of the table-level technical metadata integrity indicator = the number of non-empty table-level technical metadata / the total number of table-level technical metadata, and the score range of the table-level technical metadata integrity indicator is [0, 1];

[0080] For the table-level management metadata integrity indicator, the table original indicator data obtained by the scoring system is the total number of table-level management metadata in the first table-level metadata and the number of non-empty table-level management metadata therein, and the pre-processing manner adopted is to calculate the ratio of the number of non-empty table-level management metadata to the total number of table-level management metadata, and the calculation result is the score of the table-level management metadata integrity indicator, i.e., the score of the table-level management metadata integrity indicator = the number of non-empty table-level management metadata / the total number of table-level management metadata, and the score range of the table-level management metadata integrity indicator is [0, 1];

[0081] For the table-level business metadata integrity indicator, the table original indicator data obtained by the scoring system is the total number of table-level business metadata in the first table-level metadata and the number of non-empty table-level business metadata therein, and the pre-processing manner adopted is to calculate the ratio of the number of non-empty table-level business metadata to the total number of table-level business metadata, and the calculation result is the score of the table-level business metadata integrity indicator, i.e., the score of the table-level business metadata integrity indicator = the number of non-empty table-level business metadata / the total number of table-level business metadata, and the score range of the table-level business metadata integrity indicator is [0, 1].

[0082] D4. According to the weight of each table secondary scoring indicator and the weight of each table primary scoring indicator, the score of each table secondary scoring indicator is weighted to obtain the table score of the table level metadata of the table.

[0083] After obtaining the score of the field comprehensive scoring indicator, the score of the table level technical metadata integrity indicator, the score of the table level management metadata integrity indicator and the score of the table level business metadata integrity indicator, i.e. the score of each table secondary scoring indicator of the first table level metadata, the scoring system weights the score of each table secondary scoring indicator according to the weight of each table secondary scoring indicator and the weight of each table primary scoring indicator in the table level metadata of the first table level metadata to obtain the field score of the first table level metadata, specifically: Figure 5

[0084] In the embodiment, among the table secondary scoring indicators of the first table level metadata: the number of non-empty table level technical metadata / the total number of field level technical metadata = 0.9, so the score of the table level technical metadata integrity indicator is 0.9; the number of non-empty table level management metadata / the total number of table level management metadata = 0.75, so the score of the table level management metadata integrity indicator is 0.75; the number of non-empty table level business metadata / the total number of table level business metadata = 0.85, so the score of the table level business metadata integrity indicator is 0.85; it can be known from the above that the score of the field comprehensive scoring indicator of the first table level metadata is 0.79.

[0085] Thus, the scoring system first weights the score of each table secondary scoring indicator according to the weight of each table secondary scoring indicator to obtain the score of each table primary scoring indicator, and the calculation formula is:

[0086] The score of the table primary scoring indicator = ∑ the score of the table secondary scoring indicator * the weight of the table secondary scoring indicator;

[0087] Among the table primary scoring indicators of the first table level metadata: the score of the table level integrity indicator = the score of the table level technical metadata integrity indicator * the weight of the table level technical metadata integrity indicator + the score of the table level management metadata integrity indicator * the weight of the table level management metadata integrity indicator + the score of the table level business metadata integrity indicator * the weight of the table level business metadata integrity indicator = 0.9 * 80% + 0.75 * 10% + 0.85 * 10% = 0.88; the score of the field indicator = the score of the field comprehensive scoring indicator * the weight of the field comprehensive scoring indicator = 0.79 * 100% = 0.79.

[0088] Then, the scoring system weights the score of each table primary scoring indicator according to the weight of each table primary scoring indicator to obtain the table score of the first table level metadata, and the calculation formula is:

[0089] ​The table score of the first table level metadata = the score of the table first level score indicator * the weight of the table first level score indicator + the score of the field indicator * the weight of the field indicator = 0.88 * 60% + 0.79 * 40% = 0.844.

[0090] The table score of the first table level metadata = the score of the table first level score indicator * the weight of the table first level score indicator + the score of the field indicator * the weight of the field indicator = 0.88 * 60% + 0.79 * 40% = 0.844.

[0091] Similarly, since the table first level score indicators of each table level metadata in the embodiment are the same, and the table second level score indicators of each table level metadata are the same, and the scoring manner of each table level metadata is the same, the second table level metadata can be scored in the same manner as the first table level metadata, and the table score of the second table level metadata is, for example, 0.796.

[0092] E. The table scores of all table level metadata describing the business object are averaged to obtain the metadata comprehensive score of the business object.

[0093] In the embodiment, since the first table level metadata and the second table level metadata describe the business object, the table score of the first table level metadata and the table score of the second table level metadata can be averaged to obtain the table comprehensive score of all table level metadata contained in the business object, i.e., the metadata comprehensive score of the business object, so that the metadata comprehensive score of the business object = (the table score of the first table level metadata + the table score of the second table level metadata) / 2 = (0.844 + 0.796) / 2 = 0.82.

[0094] After obtaining the metadata comprehensive score of the business object, the metadata of the business object can be rated according to the metadata comprehensive score of the business object, and the rating basis is, for example, Figure 6The comprehensive score corresponds to the level, specifically: if the metadata comprehensive score of the business object is in the numerical range of [0.9, 1], the rating is excellent; if the metadata comprehensive score of the business object is in the numerical range of [0.8, 0.9), the rating is good; if the metadata comprehensive score of the business object is in the numerical range of [0.6, 0.8), the rating is general; if the metadata comprehensive score of the business object is in the numerical range of [0.4, 0.6), the rating is warning; if the metadata comprehensive score of the business object is in the numerical range of [0, 0.4), the rating is abnormal. In this way, the metadata comprehensive score of the business object rated as excellent, good or general (i.e. the comprehensive score is above 0.6) is high, meaning that the table-level metadata and field-level metadata used to describe the business object are reliable in quality, and the metadata comprehensive score of the business object in the present embodiment is 0.82, which is in the numerical range of [0.8, 0.9), so the metadata rating of the business object in the present embodiment is good, meaning that the table-level metadata and field-level metadata used to describe the business object are reliable in quality, and the data retrieval and knowledge graph processing result of the business object are also reliable.

[0095] In the present embodiment, each item of field-level metadata is scored first to obtain the field score of each item of field-level metadata, the field scores of all field-level metadata contained in each item of table-level metadata are then respectively averaged to obtain the field comprehensive score of all field-level metadata contained in each item of table-level metadata, and then each item of table-level metadata is scored, in which process the field comprehensive score of all field-level metadata contained in the table-level metadata is taken as the score of one of the table two-level score indicators, and the scores of the table two-level score indicators are then weighted and calculated according to the corresponding indicator weights to obtain the table score of the table-level metadata, and then the table scores of all table-level metadata are averaged to obtain the metadata comprehensive score of the business object. In this way, the metadata comprehensive score of the business object finally obtained combines the table scores of all table-level metadata and the field scores of all field-level metadata, and comprehensively considers the quality influence between the table-level metadata and the field-level metadata used to describe the business object. In this way, data retrieval and knowledge graph processing using a business object with a high comprehensive score (i.e. the comprehensive score is above 0.6 and the rating is excellent, good or general) can improve reliability.

[0096] The above description is only an embodiment of the present application, and does not limit the scope of patent protection. Those skilled in the art can make non-essential changes or substitutions based on the present application, and still fall within the scope of patent protection.

Claims

1. A method of scoring metadata of a business object, characterized by, The method comprises the following steps: A. obtaining at least one table-level metadata and at least one field-level metadata contained in each table-level metadata for describing a business object; B. scoring each field-level metadata, specifically comprising the following steps B1-B4: B1. obtaining at least two field-level first scoring indexes and at least one field-level second scoring index contained in each field-level first scoring index for scoring the field-level metadata; B2. obtaining the weight respectively given by a user to each field-level first scoring index and each field-level second scoring index; B3. obtaining field original index data of each field-level second scoring index, and calculating the score of each field-level second scoring index by pre-processing the field original index data respectively; B4. calculating the score of each field-level second scoring index by weighting according to the weight of each field-level second scoring index and the weight of each field-level first scoring index, to obtain the field score of the field-level metadata; C. calculating the average of the field score of all field-level metadata contained in each table-level metadata, to obtain the field comprehensive score of all field-level metadata contained in each table-level metadata; D. scoring each table-level metadata, specifically comprising the following steps D1-D4: D1. obtaining at least two table-level first scoring indexes and at least one table-level second scoring index contained in each table-level first scoring index for scoring the table-level metadata, and taking the field comprehensive score of all field-level metadata contained in the table-level metadata as one of the table-level second scoring indexes; D2. obtaining the weight respectively given by a user to each table-level first scoring index and each table-level second scoring index; D3. obtaining table original index data of each table-level second scoring index except the field comprehensive score, and calculating the score of each table-level second scoring index by pre-processing the table original index data respectively; D4. calculating the score of each table-level second scoring index by weighting according to the weight of each table-level second scoring index and the weight of each table-level first scoring index, to obtain the table score of the table-level metadata; E. calculating the average of the table score of all table-level metadata for describing the business object, to obtain the metadata comprehensive score of the business object.

2. The method for metadata scoring of business objects according to claim 1, characterized in that, In the step B1, the field-level first scoring index comprises a field-level integrity index, and the field-level second scoring index contained in the field-level integrity index comprises a field-level technical metadata integrity index, a field-level management metadata integrity index and a field-level business metadata integrity index.

3. The method for metadata scoring of business objects of claim 1, wherein, In the step B2, the sum of the weight of all field-level first scoring indexes of the same field-level metadata is 100%, and the sum of the weight of all field-level second scoring indexes contained in the same field-level first scoring index is 100%.

4. The method for metadata scoring of business objects of claim 1, wherein, In the step B4, the scores of the field secondary scoring indicators are weighted according to the weights of the field secondary scoring indicators to obtain the scores of the field primary scoring indicators of the field level metadata, and the scores of the field primary scoring indicators are weighted according to the weights of the field primary scoring indicators to obtain the field score of the field level metadata.

5. The method for metadata scoring of business objects of claim 1, wherein, In the step D1, the table primary scoring indicators include a table level integrity indicator and a field indicator, the table secondary scoring indicators contained in the table level integrity indicator include a table level technical metadata integrity indicator, a table level management metadata integrity indicator and a table level business metadata integrity indicator, and the field comprehensive score is taken as the table secondary scoring indicators contained in the field indicator.

6. The method for metadata scoring of business objects of claim 1, wherein, In the step D2, the sum of the weights of all the table primary scoring indicators of the same table level metadata is 100%, and the sum of the weights of all the table secondary scoring indicators contained in the same table primary scoring indicator is 100%.

7. The method for metadata scoring of business objects of claim 1, wherein, In the step D4, the scores of the table secondary scoring indicators are weighted according to the weights of the table secondary scoring indicators to obtain the scores of the table primary scoring indicators of the table level metadata, and the scores of the table primary scoring indicators are weighted according to the weights of the table primary scoring indicators to obtain the table score of the table level metadata.

8. The method for metadata scoring of business objects of claim 1, wherein, In the step E, after obtaining the metadata comprehensive score of the business object, the business object is rated according to the metadata comprehensive score of the business object.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement the steps in the metadata scoring method of the business object according to claims 1 to 8.

10. A system for metadata scoring of business objects, comprising a computer readable storage medium and a processor interconnected, characterized by, The computer readable storage medium is as claimed in claim 9.

Images