Enterprise data index management system and method based on knowledge graph

Abstract

The invention discloses an enterprise data index management system and method based on a knowledge graph, and relates to the technical field of enterprise data index management, and the method comprises the following steps: obtaining the knowledge graph of enterprise data, and converting the knowledge graph into a ternary table; calculating a data index of the enterprise based on the data in the ternary table; analyzing a normal fluctuation range and a trend of the historical data index, and analyzing the completion quality of the data index based on the normal fluctuation range and the trend; the method is used for solving the problem that an evaluation management result of an enterprise data index is prone to errors due to the fact that an existing enterprise data index management technology mostly adopts a manual judgment mode for the completion condition of the enterprise data index and a judgment threshold cannot be dynamically adjusted.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise data indicator management, and specifically to an enterprise data indicator management system and method based on a knowledge graph. Background Art

[0002] Enterprise data indicator management technology refers to the methods and tools used to collect, analyze, monitor and manage enterprise data indicators. These technologies are designed to help enterprises effectively track and evaluate their performance to support decision-making, goal setting and strategic planning.

[0003] Existing enterprise data indicator management technologies usually only perform numerical evaluation on enterprise data indicators without specific classification. The completion of enterprise data indicators is usually judged manually. The quality of enterprise data indicator completion is evaluated by manual judgment, which is highly subjective and insufficient. At the same time, enterprise data indicators have certain changing trends. If the completion of enterprise data indicators is simply judged by a fixed threshold, it is not enough to accurately evaluate. If the enterprise data indicators grow steadily, the original threshold is used to judge them at this time, and the judgment result is not accurate enough. It should be dynamically adjusted in combination with actual changes. For example, in the patent application with publication number CN114021873A, a data indicator quantification method and a smart park enterprise value evaluation system are disclosed. This solution only quantifies the data indicators and does not evaluate them. The completion of data indicators is mostly judged manually, which is highly subjective and cannot be dynamically adjusted with the changing trend of data indicators, resulting in errors in the evaluation and management results of enterprise data indicators. Existing enterprise data indicator management technologies also have the problem that the completion of enterprise data indicators is mostly judged manually and the judgment threshold cannot be dynamically adjusted, resulting in errors in the evaluation and management results of enterprise data indicators. Summary of the invention

[0004] The present invention aims to solve one of the technical problems in the prior art to at least a certain extent, by acquiring a knowledge graph of enterprise data, converting the knowledge graph into a ternary table, then searching for basic data under different data indicators, calculating data indicators based on the basic data, and then analyzing the normal fluctuation range of the data indicators through historical data indicators, and then analyzing the trend of the data indicators based on historical data indicators, and finally analyzing the completion quality of the data indicators based on the normal fluctuation range and the trend, so as to solve the problem that the existing enterprise data indicator management technology still uses manual judgment on the completion of enterprise data indicators and cannot dynamically adjust the judgment threshold, resulting in errors in the evaluation management results of enterprise data indicators.

[0005] To achieve the above objectives, in a first aspect, the present application provides an enterprise data indicator management method based on a knowledge graph, comprising the following steps:

[0006] Obtain the knowledge graph of enterprise data and convert the knowledge graph into a ternary table;

[0007] Calculate the enterprise's data indicators based on the data in the ternary table;

[0008] Analyze the normal fluctuation range and trend of historical data indicators, and analyze the completion quality of data indicators based on the normal fluctuation range and trend.

[0009] Furthermore, obtaining the knowledge graph of enterprise data and converting the knowledge graph into a ternary table includes the following sub-steps:

[0010] The ternary table includes a relationship ternary table and an attribute ternary table, the table header in the relationship ternary table includes subject, relationship and object, and the table header in the attribute ternary table includes entity, attribute and property value;

[0011] Obtain a knowledge graph of enterprise data, wherein the knowledge graph includes entities, relationships, attributes, and property values, wherein the entities and property values ​​are represented by circular nodes, and the relationships and attributes are represented by arrows;

[0012] The relationship and the attribute are directional. In the entity-relationship-entity connection, the entity in the direction of the arrow is marked as the object, and the entity in the direction of the arrow is marked as the subject; in the entity-attribute-value connection, the circular node in the direction of the arrow is the value, and the circular node in the direction of the arrow is the entity;

[0013] Based on the entity-relationship-entity connection and entity-attribute-value connection in the knowledge graph, the entities, relationships, attributes and values ​​in the knowledge graph are entered into the relationship ternary table and the attribute ternary table.

[0014] Furthermore, calculating the data indicators of the enterprise based on the data in the ternary table includes the following sub-steps:

[0015] Find basic data under different data indicators;

[0016] Calculate data indicators based on basic data.

[0017] Furthermore, finding basic data under different data indicators includes the following sub-steps:

[0018] The data indicators include financial indicators, operational indicators, market indicators and employee performance indicators;

[0019] The user associates the subject in the relational ternary table with the data indicator, and marks the correlation between the subject and the data indicator, wherein the correlation includes positive correlation and negative correlation;

[0020] Mark the entities associated with financial indicators as financial entities, the entities associated with operational indicators as operational entities, the entities associated with market indicators as market entities, and the entities associated with employee performance indicators as employee performance entities;

[0021] Find the object corresponding to the financial subject, mark it as the financial object, find the attribute value corresponding to the entity identical to the financial object in the attribute ternary table, mark it as the financial basic data;

[0022] Find the object corresponding to the operating subject, mark it as the operating object, find the attribute value corresponding to the entity that is the same as the operating object in the attribute ternary table, and mark it as the basic operating data;

[0023] Find the object corresponding to the market subject, mark it as the market object, find the property value corresponding to the entity identical to the market object in the attribute ternary table, mark it as the market basic data;

[0024] Find the object corresponding to the employee performance subject, mark it as the employee performance object, find the attribute value corresponding to the entity that is the same as the employee performance object in the attribute ternary table, and mark it as the employee performance basic data.

[0025] Furthermore, calculating data indicators based on basic data includes the following sub-steps:

[0026] Set the subscript n, n is a positive integer, and pass the symbol P in order from small to large. n It represents the financial subject, the operating subject, the market subject and the employee performance subject in turn;

[0027] P n The basic data contained in the P are numbered and represented by the symbol R(n, m), where m is a positive integer and (n, m) is the serial number of R. R(n, m) represents P n The mth basic data in;

[0028] Get the historical basic data, mark it as historical data, number the historical data of R(n, m), and use the symbol R(n, m) i Represents, where i is a positive integer and i is the sequence number of R(n, m);

[0029] For any values ​​of n and m, obtain R(n, m) i The minimum and maximum values ​​in are marked as R(n, m) respectively. min and R(n,m) max ;

[0030] For any value of n, the basic data and historical data are calculated through the indicator normalization formula to obtain the corresponding P n Data indicators;

[0031] The indicator normalization formula is configured as: Among them, K n P n Data indicators, Z>0 represents positive correlation, Z<0 represents negative correlation, and max(m) represents the maximum value of m.

[0032] Furthermore, the normal fluctuation range and trend of historical data indicators are analyzed. The completion quality of the data indicators based on the normal fluctuation range and trend analysis includes the following sub-steps:

[0033] Analyze the normal fluctuation range of data indicators through historical data indicators;

[0034] Analyze the trend of data indicators based on historical data indicators;

[0035] Analyze the completion quality of data indicators based on normal fluctuation range and trend.

[0036] Furthermore, analyzing the normal fluctuation range of data indicators through historical data indicators includes the following sub-steps:

[0037] Mark the historical data indicators as historical indicators, and n The historical indicators are marked as H n ;

[0038] For any value of n, calculate H n The mean and median of the data are calculated, and then the difference between the mean and the median is calculated, which is marked as the interval span;

[0039] Set the interval start value, which is initially zero, add the interval start value to the interval span to obtain the interval end value, thereby obtaining a range interval [interval start value, interval end value], which is marked as the traversal interval;

[0040] Starting from the interval start value of zero, gradually increase the interval start value, and the corresponding interval end value also increases synchronously, recording the H in the traversal interval n The number of , labeled as interval density;

[0041] Sort and number the interval densities from large to small, and use the symbol Q j Indicates, where j is a positive integer and j is the serial number of Q;

[0042] Mark Q1 as the fluctuation interval, starting from j=2, and judge Q jWhether there is an intersection with the fluctuation range, if so, output the interval merging signal; if not, output the polling signal;

[0043] If the output interval merges the signal, then Q j Combined with the fluctuation interval to form a new fluctuation interval, and restart the judgment again with j=2 as the starting point. If Q j If it is already in the fluctuation range, a polling signal is output; if a polling signal is output, j+1 is added and the judgment is made again;

[0044] Mark the range of the fluctuation interval as the interval size, and determine in real time whether the interval size is less than or equal to twice the interval span. When the interval size is greater than twice the interval span, stop Q j The final fluctuation range is marked as the normal fluctuation range. n There is a normal fluctuation range for each.

[0045] Furthermore, analyzing the trend of data indicators based on historical data indicators includes the following sub-steps:

[0046] For any P n , establish a plane rectangular coordinate system with time as the X-axis and historical indicators as the Y-axis, named the indicator trend analysis chart, and enter the historical indicators into the indicator trend analysis chart according to their time;

[0047] Perform polynomial regression on the indicator trend analysis chart and name the regression function as the trend;

[0048] Each P n There is a corresponding trend.

[0049] Furthermore, the completion quality of the data indicator based on the normal fluctuation range and trend analysis includes the following sub-steps:

[0050] Get the time corresponding to the data indicator currently being analyzed and name it as target time;

[0051] For any P n , substitute the target time into the trend, and name the calculated result as the estimated index point;

[0052] Determine the size relationship between the estimated index point and the normal fluctuation range. If the estimated index point is less than the minimum value of the normal fluctuation range, output an index decline signal; if the estimated index point is within the normal fluctuation range, output an index normal signal; if the estimated index point is greater than the maximum value of the normal fluctuation range, output an index rise signal;

[0053] If the output indicator drops, the estimated indicator point and the maximum value of the normal fluctuation range form a range interval, which is marked as the actual normal range; if the output indicator is normal, the normal fluctuation range is marked as the actual normal range; if the output indicator rises, the minimum value of the normal fluctuation range and the estimated indicator point form the actual normal range;

[0054] Each data indicator corresponds to a real-time normal range. The data indicator currently being analyzed is marked as the target indicator, and the target indicator is compared with the corresponding real-time normal range. If the target indicator is less than the minimum value of the real-time normal range, a low-quality indicator signal is output; if the target indicator is within the real-time normal range, a normal indicator signal is output; if the target indicator is greater than the maximum value of the real-time normal range, a high-quality indicator signal is output;

[0055] If a low-quality indicator signal is output, the completion quality of the target indicator is marked as low quality; if a normal indicator signal is output, the completion quality of the target indicator is marked as normal quality; if a high-quality indicator signal is output, the completion quality of the target indicator is marked as high quality.

[0056] In a second aspect, the present application provides an enterprise data indicator management system based on a knowledge graph, including a knowledge graph conversion module, a data indicator calculation module and a completion quality assessment module; the knowledge graph conversion module and the data indicator calculation module are respectively connected to the completion quality assessment module data;

[0057] The knowledge graph conversion module is used to obtain the knowledge graph of enterprise data and convert the knowledge graph into a ternary table;

[0058] The data indicator calculation module is used to calculate the data indicators of the enterprise based on the data in the ternary table;

[0059] The completion quality assessment module is used to analyze the normal fluctuation range and trend of historical data indicators, and analyze the completion quality of data indicators based on the normal fluctuation range and trend.

[0060] Beneficial effects of the present invention: The present invention obtains the knowledge graph of enterprise data, converts the knowledge graph into a ternary table, then searches for basic data under different data indicators, calculates data indicators based on the basic data, and then analyzes the normal fluctuation range of data indicators through historical data indicators. The advantage is that the normal fluctuation range reveals the specific grading threshold of data indicators within the enterprise, and divides them into three levels of low quality, normal quality and high quality through the normal fluctuation range, so as to intelligently evaluate the completion of enterprise data indicators, thereby improving the effectiveness and rationality of enterprise data indicator management;

[0061] The present invention analyzes the trend of data indicators based on historical data indicators, and finally analyzes the completion quality of data indicators based on normal fluctuation range and trend. The advantage is that the trend of data indicators is analyzed, and the normal fluctuation range is dynamically adjusted according to the trend, so that the completion status of enterprise data indicators can be graded more realistically, thereby improving the accuracy and rationality of enterprise data indicator management. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a functional block diagram of the system of the present invention;

[0063] Figure 2 is a schematic diagram of the knowledge graph of the present invention;

[0064] Figure 3 It is a schematic diagram of the fluctuation range of the present invention;

[0065] Figure 4 It is the indicator trend analysis diagram of the present invention;

[0066] Figure 5 The figure is a flow chart of the steps of the method of the present invention. DETAILED DESCRIPTION

[0067] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] Example 1, please refer to Figure 1 As shown, the present application provides an enterprise data indicator management system based on a knowledge graph, including a knowledge graph conversion module, a data indicator calculation module and a completion quality assessment module; the knowledge graph conversion module and the data indicator calculation module are respectively connected to the completion quality assessment module data;

[0069] The knowledge graph conversion module is used to obtain the knowledge graph of enterprise data and convert the knowledge graph into a ternary table;

[0070] The knowledge graph conversion module is configured with a knowledge graph conversion strategy, which includes:

[0071] The ternary table includes a relationship ternary table and an attribute ternary table. The table headers in the relationship ternary table include subject, relationship and object, and the table headers in the attribute ternary table include entity, attribute and property value.

[0072] See also Figure 2As shown, a knowledge graph of enterprise data is obtained. The knowledge graph includes entities, relationships, attributes, and property values. Entities and property values ​​are represented by circular nodes, and relationships and attributes are represented by arrows.

[0073] Relationships and attributes are directional. In the entity-relationship-entity connection, the entity in the direction of the arrow is marked as the object, and the entity in the direction of the arrow is marked as the subject; in the entity-attribute-value connection, the circular node in the direction of the arrow is the value, and the circular node in the direction of the arrow is the entity.

[0074] Based on the connection between entity-relationship-entity and entity-attribute-value in the knowledge graph, the entities, relationships, attributes and values ​​in the knowledge graph are entered into the relationship triple table and the attribute triple table;

[0075] In practical applications, Figure 2 Part of the data of the knowledge graph of enterprise data is shown, and part of the data of the relationship ternary table is shown in Table 1:

[0076] Table 1 Partial data of the relational ternary table

[0077] main body relation object income Include Sales revenue expenditure Include Material expenditure

[0078] Part of the data of the attribute ternary table is shown in Table 2 below:

[0079] Table 2 Partial data of attribute ternary table

[0080] entity property Value Sales revenue Numeric 2.8 million Material expenditure Numeric 1.4 million

[0081] The data indicator calculation module is used to calculate the data indicators of the enterprise based on the data in the ternary table; the data indicator calculation module includes a basic data search unit and a data indicator calculation unit;

[0082] The basic data search unit is used to search for basic data under different data indicators;

[0083] The basic data search unit is configured with a basic data search strategy, which includes:

[0084] Data indicators include financial indicators, operational indicators, market indicators, and employee performance indicators;

[0085] The user associates the subject in the relational ternary table with the data indicator, and marks the correlation between the subject and the data indicator, including positive correlation and negative correlation;

[0086] Mark the entities associated with financial indicators as financial entities, the entities associated with operational indicators as operational entities, the entities associated with market indicators as market entities, and the entities associated with employee performance indicators as employee performance entities;

[0087] In practical applications, the income and expenditure in Table 1 are both part of the financial indicators. This embodiment only takes the calculation process of the financial indicators as an example to illustrate the evaluation of the enterprise data indicators. The analysis and calculation process of the other data indicators are the same as the financial indicators. The income and expenditure are both financial entities. In the correlation relationship marked by the user, the income is positively correlated and the expenditure is negatively correlated.

[0088] Find the object corresponding to the financial subject, mark it as the financial object, find the attribute value corresponding to the entity identical to the financial object in the attribute ternary table, mark it as the financial basic data;

[0089] Find the object corresponding to the operating subject, mark it as the operating object, find the attribute value corresponding to the entity that is the same as the operating object in the attribute ternary table, and mark it as the basic operating data;

[0090] Find the object corresponding to the market subject, mark it as the market object, find the property value corresponding to the entity identical to the market object in the attribute ternary table, mark it as the market basic data;

[0091] Find the object corresponding to the employee performance subject, mark it as the employee performance object, find the attribute value corresponding to the entity identical to the employee performance object in the attribute ternary table, and mark it as the employee performance basic data;

[0092] In actual applications, in the financial subject, the object of income is sales income, and the objects of expenditure include material expenditure, advertising expenditure and wage expenditure. Through the attribute ternary table search, the value of sales income is 2.8 million, and the values ​​of material expenditure, advertising expenditure and wage expenditure are 1.4 million, 200,000 and 200,000 respectively, that is, the financial basic data are 2.8 million, 1.4 million, 200,000 and 200,000, among which 2.8 million is positively correlated, and 1.4 million, 200,000 and 200,000 are negatively correlated.

[0093] The data indicator calculation unit is used to calculate the data indicator based on the basic data;

[0094] The data indicator calculation unit is configured with a data indicator calculation strategy, which includes:

[0095] Set the subscript n, n is a positive integer, and pass the symbol P in order from small to large. n It represents the financial subject, the operating subject, the market subject and the employee performance subject in turn;

[0096] P n The basic data contained in the P are numbered and represented by the symbol R(n, m), where m is a positive integer and (n, m) is the serial number of R. R(n, m) represents P n The mth basic data in;

[0097] Get the historical basic data, mark it as historical data, number the historical data of R(n, m), and use the symbol R(n, m) i Represents, where i is a positive integer and i is the sequence number of R(n, m);

[0098] For any values ​​of n and m, obtain R(n, m) i The minimum and maximum values ​​in are marked as R(n, m) respectively. min and R(n,m) max ;

[0099] In actual applications, P1 is the financial entity. There are 4 data in the basic financial data. Therefore, 1≤m≤4, R(1,1) to R(1,4) are 2.8 million, 1.4 million, 200,000, and 200,000 respectively. For example, when m=1, R(1,1) i That is, the historical data of sales revenue in the revenue; get R(1,1) i R(1,1) min and R(1,1) max 2.1 million and 3.5 million respectively, R(1,2) i R(1,2) min and R(1,2) max 1 million and 1.6 million respectively, R(1,3) i R(1,3) min and R(1,3) max 150,000 and 250,000 respectively, R(1,4) i R(1,4) min and R(1,4) max 150,000 and 220,000 respectively;

[0100] For any value of n, the basic data and historical data are calculated through the indicator normalization formula to obtain the corresponding P n Data indicators;

[0101] The indicator normalization formula is configured as: Among them, K n P n Data indicators, Z>0 represents positive correlation, Z<0 represents negative correlation, max(m) represents the maximum value of m;

[0102] In practical applications, when n=1, the financial indicators are calculated, where m=1 is positively correlated, and m=2, 3, and 4 are negatively correlated. Therefore, the first formula is used when m=1, and the second formula is used for the remaining formulas. When summing, the corresponding formula is used according to the correlation, and then the sum is calculated; K1 is calculated to be -1.38, and the calculation result is rounded to two decimal places.

[0103] The completion quality assessment module is used to analyze the normal fluctuation range and trend of historical data indicators, and analyze the completion quality of data indicators based on the normal fluctuation range and trend; the completion quality assessment module includes a fluctuation range calculation unit, a trend analysis unit and a completion quality assessment unit;

[0104] The fluctuation range calculation unit is used to analyze the normal fluctuation range of data indicators through historical data indicators;

[0105] The fluctuation range calculation unit is configured with a fluctuation range calculation strategy, and the fluctuation range calculation strategy includes:

[0106] Mark the historical data indicators as historical indicators, and n The historical indicators are marked as H n ;

[0107] For any value of n, calculate H n The mean and median of the data are calculated, and then the difference between the mean and the median is calculated, which is marked as the interval span;

[0108] In practical applications, taking n=1 as an example, that is, taking financial indicators as an example, the historical indicator of the financial indicator is H1. Since the amount of data of the historical indicator is large, it is not convenient to show it in detail in this embodiment, so only the average and the median are given for specific analysis. The average is calculated to be -1.35, and the median is -1.22, so the interval span is 0.13; since the average and the median can both represent the normal level of the data indicator to a certain extent, a range can be set by the difference between them to obtain a reasonable interval for judging the normal fluctuation range;

[0109] Set the interval start value, which is initially zero. Add the interval start value to the interval span to get the interval end value, thereby obtaining a range interval [interval start value, interval end value], which is marked as the traversal interval;

[0110] Starting from the interval start value of zero, gradually increase the interval start value, and the corresponding interval end value also increases synchronously, recording the H in the traversal interval n The number of , labeled as interval density;

[0111] In practical applications, since the number of positively correlated basic data in the financial indicators is less than the number of negatively correlated basic data, the financial indicators are all negative numbers. Therefore, in actual use, they are usually added to a value to make them positive numbers, and the size relationship remains unchanged. In this embodiment, the financial indicator is fixedly increased by 3, that is, the financial indicator calculated in this embodiment is displayed as 3-1.38=1.62; the initial traversal interval is [0, 0.13], and the number of H1 in the traversal interval is 0 at this time, so the interval density is 0;

[0112] Sort and number the interval densities from large to small, and use the symbol Q j Indicates, where j is a positive integer and j is the serial number of Q;

[0113] See also Figure 3 As shown, Q1 is marked as the fluctuation interval, starting from j=2, and Q j Whether there is an intersection with the fluctuation range, if so, output the interval merging signal; if not, output the polling signal;

[0114] If the output interval merges the signal, then Q j Combined with the fluctuation interval to form a new fluctuation interval, and restart the judgment again with j=2 as the starting point. If Q j If it is already in the fluctuation range, a polling signal is output; if a polling signal is output, j+1 is added and the judgment is made again;

[0115] Mark the range of the fluctuation interval as the interval size, and determine in real time whether the interval size is less than or equal to twice the interval span. When the interval size is greater than twice the interval span, stop Q j The final fluctuation range is marked as the normal fluctuation range. n There is a normal fluctuation range.

[0116] In practical applications, the traversal interval with the largest interval density is used as the basis, and the fluctuation interval is gradually expanded. The traversal intervals adjacent to the fluctuation interval with large interval density are merged to Figure 3 For example, Figure 3A, B, C and D in the figure represent Q1, Q2, Q3 and Q4 respectively. The gray coordinate point is H1. Q2 and Q1 have an intersection, and the interval merging signal is output. At this time, the fluctuation interval is composed of Q1 and Q2. It is judged that the fluctuation interval is less than twice the interval span, so the merging analysis is continued. Starting with j=2, Q2 already belongs to the fluctuation interval, so Q2 is skipped. It is judged that Q3 and the fluctuation interval do not have an intersection, and a polling signal is output. Then Q4 is judged that Q has an intersection with the fluctuation interval, and an interval merging signal is output. No. At this time, the fluctuation interval includes Q1, Q2 and Q4. At this time, the fluctuation interval is less than twice the interval span, so continue to merge, starting with j=2, Q2 already belongs to the fluctuation interval, so skip Q2, judge that Q3 and the fluctuation interval have an intersection, output the interval merging signal, at this time the fluctuation interval includes Q1, Q2, Q3 and Q4, and the fluctuation interval is greater than twice the interval span, stop judging, and get the final normal fluctuation range, which is Q1, Q2, Q3 and Q4, specifically [1.51, 1.79];

[0117] The trend analysis unit is used to analyze the trend of data indicators based on historical data indicators;

[0118] The trend analysis unit is configured with a trend analysis strategy, which includes:

[0119] See also Figure 4 As shown, for any P n , establish a plane rectangular coordinate system with time as the X-axis and historical indicators as the Y-axis, named the indicator trend analysis chart, and enter the historical indicators into the indicator trend analysis chart according to their time;

[0120] Perform polynomial regression on the indicator trend analysis chart and name the regression function as the trend;

[0121] Each P n There is a trend in each direction;

[0122] In practical applications, the data indicators in this embodiment are in months, and historical indicators within one year are obtained forward, that is, historical indicators of each month in the previous 12 months are obtained. The time in the indicator trend analysis chart is replaced by numbers. For example, if the current month is May 2024, historical indicators from April 2023 to April 2024 are obtained forward, where the historical indicators from April 2023 to April 2024 are represented as 1 to 12 in sequence, and the X-axis of the data indicator of the current month is 13. The indicator trend analysis chart is constructed as follows: Figure 4 As shown, the trend obtained by polynomial regression is Y = -0.004 × X 2 +0.0994×X+1.1011, where Y is the historical indicator and X is time;

[0123] The completion quality assessment unit is used to analyze the completion quality of data indicators based on the normal fluctuation range and trend;

[0124] The completion quality assessment unit is configured with a completion quality assessment strategy, which includes:

[0125] Get the time corresponding to the data indicator currently being analyzed and name it as target time;

[0126] For any P n , substitute the target time into the trend, and name the calculated result as the estimated index point;

[0127] Determine the size relationship between the estimated index point and the normal fluctuation range. If the estimated index point is less than the minimum value of the normal fluctuation range, output an index decline signal; if the estimated index point is within the normal fluctuation range, output an index normal signal; if the estimated index point is greater than the maximum value of the normal fluctuation range, output an index rise signal;

[0128] If the output indicator drops, the estimated indicator point and the maximum value of the normal fluctuation range form a range interval, which is marked as the actual normal range; if the output indicator is normal, the normal fluctuation range is marked as the actual normal range; if the output indicator rises, the minimum value of the normal fluctuation range and the estimated indicator point form the actual normal range;

[0129] Each data indicator corresponds to a real-time normal range. The data indicator currently being analyzed is marked as the target indicator, and the target indicator is compared with the corresponding real-time normal range. If the target indicator is less than the minimum value of the real-time normal range, a low-quality indicator signal is output; if the target indicator is within the real-time normal range, a normal indicator signal is output; if the target indicator is greater than the maximum value of the real-time normal range, a high-quality indicator signal is output;

[0130] If a low-quality indicator signal is output, the completion quality of the target indicator is marked as low quality; if a normal indicator signal is output, the completion quality of the target indicator is marked as normal quality; if a high-quality indicator signal is output, the completion quality of the target indicator is marked as high quality;

[0131] In actual application, the target time is May 2024, which is represented by 13, so substitute X=13 into Y=-0.004×X 2+0.0994×X+1.1011, the estimated index point is 1.7173, and the normal fluctuation range is [1.51, 1.79]. By comparison, the estimated index point is within the normal fluctuation range, so the output index normal signal is output, that is, the actual normal range is [1.51, 1.79]. The target index in this embodiment is the financial index, and the calculated financial index is 1.62. By comparison, the financial index is within the actual normal range, and the normal index signal is output, marking the completion quality of the financial index as normal quality.

[0132] Example 2, please refer to Figure 5 As shown, the present application provides an enterprise data indicator management method based on knowledge graph, comprising the following steps:

[0133] Step S1, obtaining the knowledge graph of enterprise data and converting the knowledge graph into a ternary table; Step S1 includes the following sub-steps:

[0134] Step S101, the ternary table includes a relationship ternary table and an attribute ternary table, the table header in the relationship ternary table includes subject, relationship and object, and the table header in the attribute ternary table includes entity, attribute and property value;

[0135] Step S102, obtaining a knowledge graph of enterprise data, wherein the knowledge graph includes entities, relationships, attributes, and property values, wherein entities and property values ​​are represented by circular nodes, and relationships and attributes are represented by arrows;

[0136] Step S103: Relationships and attributes have directionality. In the entity-relationship-entity connection, the entity in the direction of the arrow pointing upward is marked as the object, and the entity in the direction of the arrow pointing in the opposite direction is marked as the subject; in the entity-attribute-value connection, the circular node in the direction of the arrow pointing upward is the value, and the circular node in the direction of the arrow pointing in the opposite direction is the entity;

[0137] Step S104, based on the connection between entity-relationship-entity and the connection between entity-attribute-value in the knowledge graph, the entities, relationships, attributes and values ​​in the knowledge graph are entered into the relationship ternary table and the attribute ternary table;

[0138] Step S2, calculating the data indicators of the enterprise based on the data in the ternary table; Step S2 includes the following sub-steps:

[0139] Step S201, searching for basic data under different data indicators;

[0140] Step S201 includes the following sub-steps:

[0141] Step S2011, data indicators include financial indicators, operational indicators, market indicators and employee performance indicators;

[0142] Step S2012, the user associates the subject in the relationship ternary table with the data indicator, and marks the correlation between the subject and the data indicator, and the correlation includes positive correlation and negative correlation;

[0143] Step S2013, marking the subject associated with the financial indicator as a financial subject, marking the subject associated with the operation indicator as an operation subject, marking the subject associated with the market indicator as a market subject, and marking the subject associated with the employee performance indicator as an employee performance subject;

[0144] Step S2014, searching for the object corresponding to the financial subject, marking it as the financial object, searching for the attribute value corresponding to the entity identical to the financial object in the attribute ternary table, marking it as the financial basic data;

[0145] Step S2015, searching for the object corresponding to the operating subject, marking it as the operating object, searching for the attribute value corresponding to the entity identical to the operating object in the attribute ternary table, marking it as the basic operating data;

[0146] Step S2016, searching for the object corresponding to the market subject, marking it as the market object, searching for the property value corresponding to the entity identical to the market object in the attribute ternary table, marking it as the market basic data;

[0147] Step S2017, searching for the object corresponding to the employee performance subject, marking it as the employee performance object, searching for the attribute value corresponding to the entity identical to the employee performance object in the attribute ternary table, marking it as the employee performance basic data;

[0148] Step S202, calculating data indicators based on basic data;

[0149] Step S202 includes the following sub-steps:

[0150] Step S2021, set the subscript n, n is a positive integer, and the subscript n is set in ascending order by the symbol P n It represents the financial subject, the operating subject, the market subject and the employee performance subject in turn;

[0151] Step S2022: n The basic data contained in the P are numbered and represented by the symbol R(n, m), where m is a positive integer and (n, m) is the serial number of R. R(n, m) represents P n The mth basic data in;

[0152] Step S2023, obtain historical basic data, mark it as historical data, number the historical data of R(n, m), and use the symbol R(n, m) i Represents, where i is a positive integer and i is the sequence number of R(n, m);

[0153] Step S2024, for any value of n and m, obtain R(n, m) i The minimum and maximum values ​​in are marked as R(n, m) respectively. min and R(n,m) max ;

[0154] Step S2025: for any value of n, calculate the basic data and historical data through the indicator normalization formula to obtain the corresponding P n Data indicators;

[0155] Step S2026, the indicator normalization formula is configured as: Among them, K n P n Data indicators, Z>0 represents positive correlation, Z<0 represents negative correlation, max(m) represents the maximum value of m;

[0156] Step S3, analyzing the normal fluctuation range and trend of historical data indicators, and analyzing the completion quality of the data indicators based on the normal fluctuation range and trend; Step S3 includes the following sub-steps:

[0157] Step S301, analyzing the normal fluctuation range of data indicators through historical data indicators;

[0158] Step S301 includes the following sub-steps:

[0159] Step S3011, mark the historical data indicators as historical indicators, and n The historical indicators are marked as H n ;

[0160] Step S3012: Calculate H for any value of n. n The mean and median of the data are calculated, and then the difference between the mean and the median is calculated, which is marked as the interval span;

[0161] Step S3013, setting the interval start value, which is initially zero, and adding the interval start value to the interval span to obtain the interval end value, thereby obtaining a range interval [interval start value, interval end value], which is marked as a traversal interval;

[0162] Step S3014, starting from the interval start value of zero, gradually increasing the interval start value, and the corresponding interval end value is also increased synchronously, recording the H in the traversed interval n The number of , labeled as interval density;

[0163] Step S3015: sort and number the interval densities in descending order, using the symbol Q j Indicates, where j is a positive integer and j is the serial number of Q;

[0164] Step S3016, mark Q1 as a fluctuation interval, starting with j=2, and determine Q j Whether there is an intersection with the fluctuation range, if so, output the interval merging signal; if not, output the polling signal;

[0165] Step S3017: If the output interval merge signal is j Combined with the fluctuation interval to form a new fluctuation interval, and restart the judgment again with j=2 as the starting point. If Q j If it is already in the fluctuation range, a polling signal is output; if a polling signal is output, j+1 is added and the judgment is made again;

[0166] Step S3018, mark the range of the fluctuation interval as the interval size, and determine in real time whether the interval size is less than or equal to twice the interval span. When the interval size is greater than twice the interval span, stop Q j The final fluctuation range is marked as the normal fluctuation range. n There is a normal fluctuation range.

[0167] Step S302, analyzing the trend of data indicators based on historical data indicators;

[0168] Step S302 includes the following sub-steps:

[0169] Step S3021, for any P n , establish a plane rectangular coordinate system with time as the X-axis and historical indicators as the Y-axis, named the indicator trend analysis chart, and enter the historical indicators into the indicator trend analysis chart according to their time;

[0170] Step S3022, performing polynomial regression on the indicator trend analysis graph, and naming the regression function as trend;

[0171] Step S3023, each P n There is a trend in each direction;

[0172] Step S303, analyzing the completion quality of the data indicators based on the normal fluctuation range and the trend;

[0173] Step S303 includes the following sub-steps:

[0174] Step S3031, obtaining the time corresponding to the data indicator currently being analyzed, and naming it as the target time;

[0175] Step S3032, for any P n , substitute the target time into the trend and name the calculated result as the estimated index point;

[0176] Step S3033, determining the magnitude relationship between the estimated index point and the normal fluctuation range, if the estimated index point is less than the minimum value of the normal fluctuation range, outputting an index decline signal; if the estimated index point is within the normal fluctuation range, outputting an index normal signal; if the estimated index point is greater than the maximum value of the normal fluctuation range, outputting an index rise signal;

[0177] Step S3034: if the indicator decline signal is output, the estimated indicator point and the maximum value of the normal fluctuation range are combined into a range interval, which is marked as the actual normal range; if the indicator is output as a normal signal, the normal fluctuation range is marked as the actual normal range; if the indicator is output as an increase signal, the minimum value of the normal fluctuation range and the estimated indicator point are combined into the actual normal range;

[0178] Step S3035: Each data indicator corresponds to a real-time normal range. The data indicator currently being analyzed is marked as a target indicator. The target indicator is compared with the corresponding real-time normal range. If the target indicator is less than the minimum value of the real-time normal range, a low-quality indicator signal is output; if the target indicator is within the real-time normal range, a normal indicator signal is output; if the target indicator is greater than the maximum value of the real-time normal range, a high-quality indicator signal is output;

[0179] Step S3036, if a low-quality indicator signal is output, the completion quality of the target indicator is marked as low quality; if a normal indicator signal is output, the completion quality of the target indicator is marked as normal quality; if a high-quality indicator signal is output, the completion quality of the target indicator is marked as high quality.

[0180] Example 3, please refer to Figure 5 As shown, Figure 5 The structural diagram of an electronic device is illustrated, and the electronic device may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps in a method for managing enterprise data indicators based on a knowledge graph are executed to achieve the following functions: obtaining a knowledge graph of enterprise data, converting the knowledge graph into a ternary table; calculating the enterprise's data indicators based on the data in the ternary table; analyzing the normal fluctuation range and trend of historical data indicators, and analyzing the completion quality of the data indicators based on the normal fluctuation range and trend.

[0181] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0182] Embodiment 4. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned enterprise data indicator management method based on the knowledge graph are executed to achieve the following functions: obtaining the knowledge graph of enterprise data and converting the knowledge graph into a ternary table; calculating the enterprise's data indicators based on the data in the ternary table; analyzing the normal fluctuation range and trend of historical data indicators, and analyzing the completion quality of the data indicators based on the normal fluctuation range and trend.

[0183] Through the description of the above implementation methods, the embodiments of the present invention can be provided as methods, systems or computer program products. Based on such an understanding, the above technical solutions can be essentially or partly contributed to the prior art in the form of software products, which can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and include several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0184] In the embodiments provided in the present application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of systems, modules and units can be electrical, mechanical or other forms.

[0185] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for enterprise data indicator management based on knowledge graph, characterized in that: The steps include: Obtain the knowledge graph of enterprise data and convert the knowledge graph into a ternary table; Calculate the enterprise's data indicators based on the data in the ternary table; Analyze the normal fluctuation range and trend of historical data indicators, and analyze the completion quality of data indicators based on the normal fluctuation range and trend.

2. According to claim 1, a method for managing enterprise data indicators based on knowledge graph, characterized in that: Obtaining the knowledge graph of enterprise data and converting the knowledge graph into a ternary table includes the following sub-steps: The ternary table includes a relationship ternary table and an attribute ternary table, the table header in the relationship ternary table includes subject, relationship and object, and the table header in the attribute ternary table includes entity, attribute and property value; Obtain a knowledge graph of enterprise data, wherein the knowledge graph includes entities, relationships, attributes, and property values, wherein the entities and property values ​​are represented by circular nodes, and the relationships and attributes are represented by arrows; The relationship and the attribute are directional. In the entity-relationship-entity connection, the entity in the direction of the arrow is marked as the object, and the entity in the direction of the arrow is marked as the subject; in the entity-attribute-value connection, the circular node in the direction of the arrow is the value, and the circular node in the direction of the arrow is the entity; Based on the entity-relationship-entity connection and entity-attribute-value connection in the knowledge graph, the entities, relationships, attributes and values ​​in the knowledge graph are entered into the relationship ternary table and the attribute ternary table.

3. According to claim 2, a method for managing enterprise data indicators based on knowledge graph, characterized in that: Calculating the enterprise's data indicators based on the data in the ternary table includes the following sub-steps: Find basic data under different data indicators; Calculate data indicators based on basic data.

4. According to claim 3, a method for managing enterprise data indicators based on knowledge graph, characterized in that: Finding basic data under different data indicators includes the following sub-steps: The data indicators include financial indicators, operational indicators, market indicators and employee performance indicators; The user associates the subject in the relational ternary table with the data indicator, and marks the correlation between the subject and the data indicator, wherein the correlation includes positive correlation and negative correlation; Mark the entities associated with financial indicators as financial entities, the entities associated with operational indicators as operational entities, the entities associated with market indicators as market entities, and the entities associated with employee performance indicators as employee performance entities; Find the object corresponding to the financial subject, mark it as the financial object, find the attribute value corresponding to the entity identical to the financial object in the attribute ternary table, mark it as the financial basic data; Find the object corresponding to the operating subject, mark it as the operating object, find the attribute value corresponding to the entity that is the same as the operating object in the attribute ternary table, and mark it as the basic operating data; Find the object corresponding to the market subject, mark it as the market object, find the attribute value corresponding to the entity identical to the market object in the attribute ternary table, mark it as the market basic data; Find the object corresponding to the employee performance subject, mark it as the employee performance object, find the attribute value corresponding to the entity that is the same as the employee performance object in the attribute ternary table, and mark it as the employee performance basic data.

5. According to claim 4, a method for managing enterprise data indicators based on knowledge graph, characterized in that: Calculating data indicators based on basic data includes the following sub-steps: Set the subscript n, n is a positive integer, and pass the symbol P in order from small to large. n It represents the financial subject, the operating subject, the market subject and the employee performance subject in turn; P n The basic data contained in the P are numbered and represented by the symbol R(n, m), where m is a positive integer and (n, m) is the serial number of R. R(n, m) represents P n The mth basic data in; Get the historical basic data, mark it as historical data, number the historical data of R(n, m), and use the symbol R(n, m) i Represents, where i is a positive integer and i is the sequence number of R(n, m); For any values ​​of n and m, obtain R(n, m) i The minimum and maximum values ​​in are marked as R(n, m) respectively. min and R(n,m) max ; For any value of n, the basic data and historical data are calculated through the indicator normalization formula to obtain the corresponding P n Data indicators; The indicator normalization formula is configured as: Among them, K n P n Data indicators, Z>0 represents positive correlation, Z<0 represents negative correlation, and max(m) represents the maximum value of m.

6. The enterprise data indicator management method based on knowledge graph according to claim 5 is characterized in that: Analyze the normal fluctuation range and trend of historical data indicators. The completion quality of data indicators based on the normal fluctuation range and trend analysis includes the following sub-steps: Analyze the normal fluctuation range of data indicators through historical data indicators; Analyze the trend of data indicators based on historical data indicators; Analyze the completion quality of data indicators based on normal fluctuation range and trend.

7. The enterprise data indicator management method based on knowledge graph according to claim 6 is characterized in that: Analyzing the normal fluctuation range of data indicators through historical data indicators includes the following sub-steps: Mark the historical data indicators as historical indicators, and n The historical indicators are marked as H n ; For any value of n, calculate H n The mean and median of the data are calculated, and then the difference between the mean and the median is calculated, which is marked as the interval span; Set the interval start value, which is initially zero, add the interval start value to the interval span to obtain the interval end value, thereby obtaining a range interval [interval start value, interval end value], which is marked as the traversal interval; Starting from the interval start value of zero, gradually increase the interval start value, and the corresponding interval end value also increases synchronously, recording the H in the traversal interval n The number of , labeled as interval density; Sort and number the interval densities from large to small, and use the symbol Q j Indicates, where j is a positive integer and j is the serial number of Q; Mark Q1 as the fluctuation interval, starting from j=2, and judge Q j Whether there is an intersection with the fluctuation range, if so, output the interval merging signal; if not, output the polling signal; If the output interval merges the signal, then Q j Combined with the fluctuation interval to form a new fluctuation interval, and restart the judgment again with j=2 as the starting point. If Q j If it is already in the fluctuation range, a polling signal is output; if a polling signal is output, j+1 is added and the judgment is made again; Mark the range of the fluctuation interval as the interval size, and determine in real time whether the interval size is less than or equal to twice the interval span. When the interval size is greater than twice the interval span, stop Q j The final fluctuation range is marked as the normal fluctuation range. n There is a normal fluctuation range for each.

8. The enterprise data indicator management method based on knowledge graph according to claim 7 is characterized in that: Analyzing the trend of data indicators based on historical data indicators includes the following sub-steps: For any P n , establish a plane rectangular coordinate system with time as the X-axis and historical indicators as the Y-axis, named the indicator trend analysis chart, and enter the historical indicators into the indicator trend analysis chart according to their time; Perform polynomial regression on the indicator trend analysis chart and name the regression function as the trend; Each P n There is a corresponding trend.

9. The enterprise data indicator management method based on knowledge graph according to claim 8 is characterized in that: The quality of data indicators based on normal fluctuation range and trend analysis includes the following sub-steps: Get the time corresponding to the data indicator currently being analyzed and name it as target time; For any P n , substitute the target time into the trend, and name the calculated result as the estimated index point; Determine the relationship between the estimated index point and the normal fluctuation range. If the estimated index point is less than the minimum value of the normal fluctuation range, output the indicator decline signal. If the estimated indicator point is within the normal fluctuation range, the indicator normal signal is output; If the estimated indicator point is greater than the maximum value of the normal fluctuation range, an indicator rising signal is output; If the output indicator declines, the estimated indicator point and the maximum value of the normal fluctuation range form a range interval, which is marked as the actual normal range; If the output indicator is a normal signal, the normal fluctuation range is marked as the actual normal range; If the output indicator rises, the minimum value of the normal fluctuation range and the estimated indicator point form the actual normal range; Each data indicator corresponds to a real-time normal range. The data indicator currently being analyzed is marked as the target indicator, and the target indicator is compared with the corresponding real-time normal range. If the target indicator is less than the minimum value of the real-time normal range, a low-quality indicator signal is output; if the target indicator is within the real-time normal range, a normal indicator signal is output; if the target indicator is greater than the maximum value of the real-time normal range, a high-quality indicator signal is output; If a low-quality indicator signal is output, the completion quality of the target indicator is marked as low quality; If a normal indicator signal is output, the completion quality of the target indicator is marked as normal quality; if a high-quality indicator signal is output, the completion quality of the target indicator is marked as high quality.

10. An enterprise data indicator management system based on knowledge graph, used to implement an enterprise data indicator management method based on knowledge graph according to any one of claims 1 to 9, characterized in that: It includes a knowledge graph conversion module, a data indicator calculation module and a completion quality assessment module; the knowledge graph conversion module and the data indicator calculation module are respectively connected to the completion quality assessment module data; The knowledge graph conversion module is used to obtain the knowledge graph of enterprise data and convert the knowledge graph into a ternary table; The data indicator calculation module is used to calculate the data indicators of the enterprise based on the data in the ternary table; The completion quality assessment module is used to analyze the normal fluctuation range and trend of historical data indicators, and analyze the completion quality of data indicators based on the normal fluctuation range and trend.

Images