JP6552882B2 - Multidimensional data analysis support device - Google Patents

Description

  The present invention relates to a multidimensional data analysis support apparatus that supports analysis of multidimensional data consisting of a large number of data items.

  Conventionally, for the purpose of quality evaluation, data is analyzed using a computer, and a report in the form of a form showing the analysis result has been created. Such a report plays an important role as a unified index for various business improvements, but it takes a lot of time to create a report. Thus, attempts have been made to automate the creation of such reports (see, for example, Patent Document 1).

  Specifically, in the prior art described in Patent Document 1, observed data is analyzed according to a predetermined analysis procedure, and a table and a chart in which the analysis result is shown are arranged in a specified layout. Is configured to create automatically.

Japanese Patent No. 5556164

However, the prior art has the following problems.
In the prior art described in Patent Document 1, a report showing the result of routine analysis by a predetermined analysis procedure is automatically created. Therefore, when analyzing multidimensional data, the method described in Patent Document 1 is used. Even when the conventional technique is applied, only a typical analysis result can be obtained.

  In addition, in order to find trends or problems in multidimensional data, when performing atypical analysis such as analysis from different viewpoints or analysis that digs deep under specific conditions, from typical analysis results, Humans need to determine empirically whether the need for additional analysis or the conditions for additional analysis are. Therefore, there is a problem that it is difficult to perform atypical analysis for a person with little analysis experience.

  The present invention has been made to solve the problems as described above, and it is an object of the present invention to provide a multidimensional data analysis support device that realizes easy atypical analysis of multidimensional data.

The multidimensional data analysis support device according to the present invention includes an analysis input data, a recommended axis item, an x axis item, a y axis item, and a y axis item for each x axis item and recommended axis item from the analysis input data. A multidimensional data analysis support apparatus to which a comparison condition in which a y-axis intermediate value and a y-axis aggregation formula for calculating a value are defined is input, and the comparison condition and recommended axis items are input from the analysis input data According to the recommended axis-specific intermediate tabulation unit for calculating the y-axis intermediate value for each x-axis item and the recommended axis item, and the y-axis intermediate value for each of the x-axis item and the recommended axis item calculated by the recommended axis-specific intermediate tabulation unit. A recommendation totaling unit that calculates the value of the y-axis item for each x-axis item and the recommended axis item as a y-axis recommended total value using a y-axis totaling formula, and the x-axis item and the recommended axis calculated by the recommendation totaling unit Item y-axis recommended tally value The correlation of the x-axis item and a y-axis recommendation aggregate value, and a data output unit that outputs for each recommendation axis items, comprising a comparison condition, x-axis items and tendency of the y-axis recommendation aggregate value for each recommendation axis items A recommendation method in which parameters for analyzing the parameters are set is further defined, and the recommendation totaling unit determines the recommended method as a parameter according to the recommendation method from the calculated x-axis item and the y-axis recommended total value for each recommended axis item. Using a statistical formula for quantifying the set tendency, a recommendation evaluation value is further calculated for each recommended axis item, and the data output unit calculates the recommended evaluation value for each recommended axis item calculated by the recommendation aggregation unit. It is further output .

  According to the present invention, the x-axis item and the recommended axis are calculated by calculating the y-axis recommendation total value for each x-axis item and the recommended axis item according to the set comparison condition and the recommended axis item from the input data for analysis. Using the y-axis recommended total value for each item, the correlation between the x-axis item and the y-axis recommended total value is output for each recommended axis item. As a result, it is possible to obtain a multidimensional data analysis support device that realizes the atypical analysis facilitation for multidimensional data.

It is a block diagram which shows the structure of the multidimensional data analysis assistance apparatus in Embodiment 1 of this invention. It is explanatory drawing which shows an example of the table of failure information DB of FIG. It is explanatory drawing which shows an example of the table of contract information DB of FIG. It is explanatory drawing which shows an example of the comparison conditions of FIG. It is explanatory drawing which shows an example of the recommendation axis | shaft item of FIG. It is explanatory drawing which shows an example of the recommendation system of FIG. It is explanatory drawing which shows an example of the table of intermediate | middle total value DB of FIG. It is explanatory drawing which shows an example of the table of integrated total value DB of FIG. It is explanatory drawing which shows an example of the table of recommendation total value DB of FIG. It is explanatory drawing which shows an example of the table of recommendation evaluation value DB of FIG. It is explanatory drawing which shows an example of the main quality report of FIG. It is explanatory drawing which shows an example of the recommendation quality report of FIG. It is explanatory drawing which shows another example of the recommendation quality report of FIG. It is a flowchart which shows a series of operation examples of the multidimensional data analysis assistance apparatus in Embodiment 1 of this invention.

  Hereinafter, a multidimensional data analysis support device according to the present invention will be described according to a preferred embodiment with reference to the drawings. In the description of the drawings, the same or corresponding portions are denoted by the same reference numerals, to omit redundant description. Further, in the embodiment, a case where the present invention is applied to multi-dimensional data related to maintenance work of a maintenance target device (for example, elevator) will be exemplified.

Embodiment 1
FIG. 1 is a block diagram showing a configuration of a multidimensional data analysis support apparatus 1 according to Embodiment 1 of the present invention. Here, the multidimensional data analysis support device 1 according to the first embodiment analyzes the input input data for analysis according to the input comparison condition 4, the recommended axis item 5 and the recommendation method 6, and Output analysis results.

  The multidimensional data analysis support device 1 is realized by, for example, a CPU that executes a program stored in a memory and a processing circuit such as a system LSI. Each database (DB) in which various data is stored is stored in a memory.

  First, each input element to the multidimensional data analysis support device 1 will be described. Here, a case where data included in each of the failure information DB 2 and the contract information DB 3 is input to the multidimensional data analysis support device 1 as the analysis input data to be analyzed is exemplified.

  FIG. 2 is an explanatory view showing an example of a table of the failure information DB 2 of FIG. As shown in FIG. 2, the failure information DB 2 includes a table regarding failure information of the maintenance target device. Specifically, the table in FIG. 2 associates failure ID, contract ID, failure time, and classification item X as data items.

  In FIG. 2, the fault ID indicates an identifier for identifying a fault. The contract ID indicates an identifier for identifying the content of the maintenance contract for maintaining the maintenance target device in which the failure corresponding to the failure ID has occurred. The failure time indicates the date on which the failure corresponding to the failure ID occurred. The classification item X indicates what kind of failure the fault ID corresponds to.

  In FIG. 2, for example, in the data on the first line, the failure ID is failure 0001, the content of the maintenance contract corresponding to the failure ID is contract 00002, and the failure time corresponding to the failure ID is August 2014. This indicates that the failure type corresponding to the failure ID is class X01.

  FIG. 3 is an explanatory diagram showing an example of the table of the contract information DB 3 in FIG. As shown in FIG. 3, the contract information DB 3 includes a table regarding maintenance contract information of the maintenance target device. Specifically, the table of FIG. 3 associates a contract ID, a contract time, a model, and a classification item Y as data items.

  In FIG. 3, the contract ID indicates an identifier for identifying the content of the maintenance contract of the maintenance target device. The contract time indicates the date when the maintenance contract corresponding to the contract ID is concluded. The model indicates the model of the maintenance target device corresponding to the contract ID. The classification item Y indicates what kind of maintenance contract corresponding to the contract ID is classified into.

  In FIG. 3, for example, in the first line of data, the contract ID is contract 00001, the maintenance contract corresponding to the contract ID is April 1, 2010, and the maintenance target corresponding to the contract ID is The model of the device is model 1, and the type of maintenance contract corresponding to the contract ID is class Y02.

  Moreover, since each table of FIG. 2 and FIG. 3 includes data items of contract ID, the respective tables are associated with each other.

  As described above, the multidimensional data analysis support apparatus 1 receives the multidimensional data related to the maintenance work of the maintenance target device as shown in FIGS. 2 and 3 as the input data for analysis.

  FIG. 4 is an explanatory view showing an example of the comparison condition 4 of FIG. As shown in FIG. 4, the comparison condition 4 includes the x-axis item, the y-axis item, the y-axis aggregation formula, the first y-axis intermediate value, the first composite function, the second y-axis intermediate value, and the second A composite function and a narrowing condition are defined. Each item of the comparison condition 4 is set by the user.

  In FIG. 4, in each item of the x-axis item and the y-axis item, parameters for analyzing the correlation between two parameters derivable from the input data for analysis are set. Here, to analyze the correlation between the failure month and the failure rate from the analysis input data, the failure month is set as the x-axis item and the failure rate is set as the y-axis item. ing.

  The items of the first y-axis intermediate value and the second y-axis intermediate value are parameters necessary to calculate the value of the y-axis item for each of the x-axis items and the recommended axis item 5 described later from the input data for analysis Is set. Here, it is necessary to calculate the failure rate for each failure month and model from the data included in each of the failure information DB 2 and the contract information DB 3.

  Therefore, 2 of the first y-axis intermediate value and the second y-axis intermediate value as the y-axis intermediate value for calculating the value of the y-axis item for each x-axis item and recommended axis item 5 from the input data for analysis Two parameters are set. Specifically, the case where the number of failure information DB2 is set as the first y-axis intermediate value and the number of contract information DB3 is set as the second y-axis intermediate value is illustrated.

  In the item of the y-axis aggregation formula, a function necessary to calculate the value of the y-axis item for each of the x-axis item and the recommended axis item 5 from the input data for analysis is set. Here, from the first y-axis intermediate value and the second y-axis intermediate value, a function for calculating the value of the y-axis item for each x-axis item and recommended axis item 5 is set as the y-axis aggregation expression . Specifically, the case where [first y-axis intermediate value] / [second y-axis intermediate value] is set as a function of the y-axis aggregation formula is illustrated.

  In the item of the first combination function, a function for combining the first y-axis intermediate value for each of the x-axis item and the recommendation axis item 5 for each same value of the x-axis item is set. Here, a case where an addition function is set as the first synthesis function is illustrated.

  In the item of the second composite function, a function for combining the second y-axis intermediate value for each of the x-axis item and the recommended-axis item 5 with the same value of the x-axis item is set. Here, a case where an addition function is set as the second synthesis function is illustrated.

  A condition for narrowing down the number of input data for analysis is set as the narrowing condition. Here, in the data included in the failure information DB 2, a condition to be analyzed using data in which the classification item X belongs to the classification X 01 is set as the narrowing-down condition. In this case, a fault belonging to the class X01 is analyzed.

  FIG. 5 is an explanatory diagram showing an example of the recommended axis item 5 in FIG. FIG. 6 is an explanatory view showing an example of the recommendation method 6 of FIG. The items of the recommendation axis item 5 and the recommendation scheme 6 are set by the user.

  In FIGS. 5 and 6, each item of recommendation axis item 5 and recommendation method 6 is specified based on the correlation between two parameters set in each item of x axis item and y axis item of comparison condition 4 Parameters for further analysis of the tendency are set. It is also possible to set a plurality of recommendation axis items 5 instead of one. In this example, assuming that the user wants to know a model whose failure is increasing, the model is set as the recommended axis item 5 and the increasing tendency is set as the recommendation method 6. . In this case, the increasing tendency of the failure rate with respect to the failure month is quantitatively analyzed for each model of the maintenance target device.

  As described above, the multidimensional data analysis support device 1 includes the comparison condition 4, the recommended axis item 5 and the recommendation method 6 corresponding to the conditions for analyzing the input data for analysis as shown in FIGS. Is also input together with the input data for analysis.

  Note that various parameters can be set for each of the x-axis item, the y-axis item, the recommended axis item 5 and the recommendation method 6 according to the contents of the analysis input data.

  As the x-axis item, for example, a year, a branch office, or the like can be set. As the y-axis item, for example, repair work time, repair cost, or failure stop time (MTTR) can be set. As the recommended axis item 5, for example, a contract form, a failure part, a failure phenomenon, a failure cause, a facility application, an installation environment, a management department, or an installation year can be set. As the recommendation method 6, for example, it is possible to set a decreasing tendency, a magnitude trend of an average value, a steep rising trend, a steep falling trend, or a seasonal fluctuation tendency.

  In the first embodiment, as an example of the combination of the parameter that can be set in the recommendation axis item 5 and the parameter that can be set in the recommendation method 6, the case of combining the model and the increase tendency is illustrated. It goes without saying that various combinations can be set as a combination of parameters that can be set in the recommendation axis item 5 and parameters that can be set in the recommendation scheme 6. For example, when the contract form is set to the recommendation axis item 5 and the decreasing tendency is set to the recommendation method 6, the decreasing tendency of the failure rate with respect to the failure occurrence month is quantitatively analyzed for each contract form of the maintenance target device Become.

  The first y-axis intermediate value in the comparison condition 4 and the second y-axis intermediate value according to the contents of the analysis input data and the contents of the parameters set in the x-axis item, the y-axis item and the recommended axis item 5 It goes without saying that the setting contents of the value, the y-axis aggregation formula, the first combination function, the second combination function, and the narrowing-down conditions are appropriately changed.

  Next, each component of the multidimensional data analysis support device 1 will be described. In FIG. 1, the multidimensional data analysis support device 1 includes a recommended axis-specific intermediate totaling unit 11, an intermediate totaling value DB 12, an integrated totaling unit 13, a recommended totaling unit 14, an integrated totaling value DB 15, a recommended totaling value DB 16, and a recommended evaluation value DB 17. And a data output unit 18.

  The recommended axis-by-recommendation-basis intermediate tabulation unit 11 sets the first for each x axis item and recommended axis item 5 according to the comparison condition 4 and the recommended axis item 5 from each of the failure information DB 2 and the contract information DB 3 input as analysis input data The y-axis intermediate value and the second y-axis intermediate value are calculated, and the calculation result is stored in the intermediate total value DB 12.

  Specifically, the recommended axis-specific intermediate totaling unit 11 associates each of the failure information DB 2 and the contract information DB 3 with a contract ID, and sets the failure occurrence month set as the x-axis item and the model set as the recommended axis item 5. The number of failures belonging to the category X01 counted in is taken as a first y-axis intermediate value. Further, the recommendation axis-based intermediate totalization unit 11 sets the number of maintenance contracts counted for each model from the contract information DB 3 as the second y-axis intermediate value.

  Here, the intermediate summary value DB 12 will be described. FIG. 7 is an explanatory diagram showing an example of a table of the intermediate summary value DB 12 of FIG. As shown in FIG. 7, the intermediate total value DB 12 includes a table related to the calculation result by the recommended axis-specific intermediate total unit 11. Specifically, the table in FIG. 7 associates the x-axis item, the recommended axis item 5, the first y-axis intermediate value, and the second y-axis intermediate value as data items.

  In FIG. 7, for example, the first line of data indicates that the number of failures occurring in model 1 in January is 35, and the number of maintenance contracts for model 1 is 5001.

  Returning to the description of FIG. 1, the integrated tabulation unit 13 determines the x-axis item stored in the intermediate totalized value DB 12 and the first y-axis intermediate value and the second y-axis intermediate value for each recommended axis item 5 The value of the y-axis item is calculated for each axis item as a y-axis integrated tally value using the first combination function, the second combination function, and the y-axis tally equation, and the calculation result is stored in the integrated tally value DB 15.

  Specifically, the first addition value is calculated by adding the first y-axis intermediate value for each failure occurrence month and each model using the addition function for each failure occurrence month. Further, the second added value is calculated by adding the second y-axis intermediate value for each failure occurrence month and each model using the addition function for each failure occurrence month. Further, a value obtained by dividing the first addition value by the second addition value for each failure occurrence month is set as a y-axis integrated total value.

  Here, the integrated total value DB 15 will be described. FIG. 8 is an explanatory diagram showing an example of a table of the integrated total value DB 15 of FIG. As shown in FIG. 8, the integrated totalized value DB 15 includes a table regarding the calculation result by the integrated totalizing unit 13. Specifically, the table in FIG. 8 associates the x-axis item and the y-axis integrated total value as data items.

  In FIG. 8, for example, data in the first row indicate that the failure rate occurring in all models is 0.0055 when the month in which the failure occurs is January. Moreover, this numerical value calculates the 1st addition value (= 35 + 36 + 36 + 37) and the 2nd addition value (= 5001 + 6009 + 808 2 + 7282) corresponding to January from middle total value DB12 shown in FIG. It is a value obtained by dividing by the second addition value.

  Returning to the description of FIG. 1, the recommendation aggregation unit 14 determines the x-axis item stored in the intermediate aggregation value DB 12 and the first y-axis intermediate value and the second y-axis intermediate value for each of the recommended axis items 5 The value of the y-axis item is calculated as the y-axis recommendation aggregation value using the y-axis aggregation formula for each of the axis item and the recommendation axis item 5, and the calculation result is stored in the recommendation aggregation value DB 16. Specifically, for each failure month and model, a value obtained by dividing the first y-axis intermediate value by the second y-axis intermediate value is taken as the y-axis recommended aggregation value.

  Here, the recommended total value DB 16 will be described. FIG. 9 is an explanatory view showing an example of a table of the recommendation total value DB 16 of FIG. As shown in FIG. 9, the recommendation total value DB 16 includes a table regarding the calculation result of the y-axis recommendation total value by the recommendation totaling unit 14. Specifically, the table in FIG. 9 associates the x-axis item, the recommended axis item 5 and the y-axis recommendation total value as data items.

  In FIG. 9, for example, the data in the first line indicates that the failure rate generated in the model 1 when the failure month is January is 0.0070. Further, this numerical value is calculated using the first y-axis intermediate value (= 35) and the second y-axis intermediate value (= 5001) corresponding to January and model 1 from the intermediate aggregated value DB12 shown in FIG. 7. , A value obtained by dividing the first y-axis intermediate value by the second y-axis intermediate value.

  Returning to the description of FIG. 1, the recommendation totaling unit 14 further calculates a recommended evaluation value for each recommended axis item 5 according to the recommendation method 6 from the calculated x-axis item and the y-axis recommended total value for each recommended axis item 5. Then, the calculation result is stored in the recommended evaluation value DB 17. Specifically, a recommendation evaluation value is calculated for each model by using a statistical formula corresponding to the recommendation method 6.

  Here, since the recommendation method 6 is set as the increasing tendency, in order to quantify the increasing tendency of the failure rate for each model, for example, the following equation (1) is used as a statistical formula corresponding to the increasing tendency. Can be used.

However, in equation (1), A _i is the y-axis recommended total value for i month of model N, B _i is an integer i (a series increasing by 1), and {A _i } and {B _i } (both The correlation coefficient R of length n = 12) is used as the recommended evaluation value.

  Further, the recommendation aggregation unit 14 further calculates the recommendation order of the recommendation axis item from the recommendation evaluation value for each of the calculated recommendation axis items 5, and stores the calculation result in the recommendation evaluation value DB 17. Note that the recommendation order is a ranking of the degree of tendency set by the recommendation method 6, and the higher the degree, the higher the order.

  For example, as described above, when the recommendation evaluation value is calculated for each model using Equation (1), the tendency to increase the failure rate becomes higher as the recommendation evaluation value is larger. In this case, ranking is performed based on the magnitude relationship of the recommended evaluation values for each model, and the recommendation ranking of the model having the largest recommended evaluation value is set to be first.

  Here, the recommended evaluation value DB 17 will be described. FIG. 10 is an explanatory diagram showing an example of a table of the recommended evaluation value DB 17 of FIG. As shown in FIG. 10, the recommendation evaluation value DB 17 includes a table relating to the calculation result of the recommendation evaluation value and the recommendation order by the recommendation totaling unit 14. Specifically, the table of FIG. 10 associates the recommended axis item 5, the recommended evaluation value, and the recommendation rank as data items.

  In FIG. 10, for example, the data in the first row indicates that the recommendation evaluation value of model 1 is 0.84 and the recommendation ranking is first. That is, the model 1 with the first recommendation rank indicates that the failure rate has the highest increasing tendency among the models 1 to 4. Further, this numerical value is a value calculated according to the equation (1) using the y-axis recommendation total value of each month from January to December corresponding to the model 1 from the recommendation total value DB 16 shown in FIG. .

  The data output unit 18 outputs the correlation between the x-axis item and the y-axis integrated tally as main quality report 7 using the y-axis integrated tally for each x-axis item stored in the integrated tally DB 15. . Specifically, the data output unit 18 outputs the correlation between the month of failure occurrence and the failure rate of all models as the main quality report 7.

  Here, the main quality report 7 will be described. FIG. 11 is an explanatory view showing an example of the main quality report 7 of FIG. As shown in FIG. 11, the main quality report 7 is output in a form in which the y-axis integrated total value for each failure occurrence month is plotted in a graph with the vertical axis representing the failure rate and the horizontal axis representing the failure occurrence month. is there. That is, in the main quality report 7, the correlation between the failure occurrence month and the failure rate of all the models is shown by a graph.

  Further, the data output unit 18 uses the x-axis item and the y-axis recommended total value for each recommended axis item 5 stored in the recommended total value DB 16 to determine the correlation between the x-axis item and the y-axis recommended total value. Each is output as a recommended quality report 8. Specifically, the data output unit 18 outputs the correlation between the failure occurrence month and the failure rate of the model as a recommended quality report 8 for each model.

  Furthermore, the data output unit 18 also outputs a recommendation order for each recommended axis item 5 stored in the recommended evaluation value DB 17. Specifically, the data output unit 18 also outputs a recommendation order for each model.

  Here, the recommended quality report 8 will be described. FIG. 12 is an explanatory diagram showing an example of the recommended quality report 8 of FIG. FIG. 13 is an explanatory diagram showing another example of the recommended quality report 8 of FIG. As shown in FIG. 12 and FIG. 13, the recommended quality report 8 shows the y-axis recommended aggregate value for each failure occurrence month plotted on a graph with the failure rate on the vertical axis and the failure occurrence month on the horizontal axis. Is output. That is, in the recommended quality report 8, the correlation between the failure occurrence month and the failure rate of the model is shown for each model in a graph.

  Further, as shown in FIG. 12 and FIG. 13, the recommendation order is also shown in the graph of each model. That is, the graph corresponding to model 1 shown in FIG. 12 shows that the recommendation order is first, and the graph corresponding to model 3 shown in FIG. 13 shows that the recommendation order is second Has been.

  With this configuration, the user can check the correlation between the two parameters set in the x-axis item and the y-axis item of the comparison condition 4. Further, the correlation between the x-axis item and the y-axis item can be confirmed for each data item set in the recommendation axis item 5. Furthermore, the specific tendency indicated by the correlation between the x-axis item and the y-axis item obtained for each of the recommended axis items 5 can be quantitatively confirmed according to the content set in the recommendation system.

  12 and 13 illustrate the case where the recommendation order corresponding to each model is shown in order to quantitatively indicate the increasing tendency of the failure rate of each model to the user. The recommendation evaluation value corresponding to may be indicated.

  Although the main quality report 7 and the recommendation quality report 8 illustrate the case where each correlation is represented in the form of a graph, each correlation may be represented in any format. Also, the data output unit 18 may output any of the contents of the main quality report 7 and the recommendation quality report 8 as long as the user can confirm the contents of the main quality report 7 and the recommendation quality report 8. For example, You may comprise so that it may display on the screen of a display apparatus.

  Next, a series of operation examples of the multidimensional data analysis support apparatus 1 will be described with reference to FIG. FIG. 14 is a flowchart showing a series of operation examples of the multidimensional data analysis support device 1 according to the first embodiment of the present invention.

  In step S101, the recommended axis-by-axis intermediate totaling unit 11 determines each x-axis item and recommended axis item 5 according to the comparison condition 4 and the recommended axis item 5 from the input data for analysis (here, the failure information DB2 and the contract information DB3). The y-axis intermediate value is calculated to be stored in the intermediate totalized value DB 12.

  In step S102, the integrated tabulation unit 13 calculates the value of the y-axis item for each x-axis item from the intermediate tabulation value DB 12 as the y-axis integrated tabulation value using the y-axis tabulation formula, and stores the calculated value in the integrated tabulation value DB 15. To do.

  In step S103, the recommendation totaling unit 14 calculates the value of the y-axis item for each x-axis item and the recommended axis item 5 as the y-axis recommended total value using the y-axis totaling formula from the intermediate total value DB 12, and recommends it. Stored in the total value DB 16.

  In step S <b> 104, the recommendation totaling unit 14 calculates a recommended evaluation value for each recommended axis item 5 from the recommended total value DB 16 according to the recommendation method 6, and calculates the recommended axis item 5 from the recommended evaluation value for each recommended axis item 5. The recommendation order is calculated, and these calculation results are stored in the recommendation evaluation value DB 17.

  In step S105, the data output unit 18 creates and outputs the main quality report 7 from the integrated aggregation value DB 15, and creates and outputs the recommendation quality report 8 from the recommendation aggregation value DB 16 and the recommendation evaluation value DB 17.

  As described above, the main quality report 7 and the recommendation quality report 8 are output by the multidimensional data analysis support device 1 executing the respective processes of step S101 to step S105. A desired analysis can be easily performed on the input data for analysis simply by setting the comparison condition 4, the recommendation axis item 5 and the recommendation method 6 according to the content.

  As described above, according to the first embodiment, the recommendation axis classified intermediate aggregation unit calculates the y axis intermediate value for each x axis item and recommended axis item according to the comparison condition and the recommended axis item from the input data for analysis, and x A recommendation totaling unit that calculates a value of a y-axis item for each x-axis item and each recommended axis item from a y-axis intermediate value for each axis item and each recommended axis item as a y-axis recommended total value using a y-axis totaling formula; The data output unit is configured to output, for each recommended axis item, a correlation between the x axis item and the y axis recommended total value from the y axis recommended total value for each x axis item and recommended axis item.

  This makes it possible to facilitate atypical analysis of multidimensional data. In addition, since the analysis result corresponding to the input by the user is output without depending on the analysis skill of the user, the unexpected notice is given to the user regarding multi-dimensional data, and it is atypical even for the user with few analysis skills. The result of a typical analysis. Furthermore, by presenting characteristic analysis cases to the user, such analysis cases are recognized as preliminary knowledge in maintenance work, and as a result, maintenance quality is improved.

  Although Embodiment 1 exemplifies the case where the present invention is applied to multi-dimensional data relating to maintenance work of a maintenance target device (e.g., elevator), the present invention is not limited to this, and any multi-dimensional data Needless to say, this is also applicable to the above.

  1 multi-dimensional data analysis support device, 2 failure information DB, 3 contract information DB, 4 comparison condition, 5 recommendation axis item, 6 recommendation method, 7 main quality report, 8 recommendation quality report, 11 recommendation axis classified total part, 12 Intermediate total value DB, 13 integrated total unit, 14 recommended total unit, 15 integrated total value DB, 16 recommended total value DB, 17 recommended evaluation value DB, 18 data output unit.

Claims (3)

Input data for analysis,
Recommendation axis item,
The x-axis item, y-axis item, and y-axis intermediate value and y-axis aggregation formula for calculating the value of the y-axis item for each of the x-axis item and the recommended axis item from the analysis input data are defined. And the compared conditions
Is a multidimensional data analysis support device,
A recommended axis classified intermediate aggregation unit that calculates the y axis intermediate value for each of the x axis item and the recommended axis item according to the comparison condition and the recommended axis item from the analysis input data;
The value of the y-axis item for each of the x-axis item and the recommended axis item is calculated from the x-axis item and the y-axis intermediate value for each recommended axis item calculated by the recommended axis-specific intermediate totaling unit. A recommendation counting unit that calculates a y-axis recommended total value using an axis totaling formula;
A correlation between the x-axis item and the y-axis recommended total value is output for each recommended axis item from the x-axis item and the y-axis recommended total value for each recommended axis item calculated by the recommendation totaling unit. Data output unit, and
Equipped with a,
The comparison condition further defines a recommendation method in which a parameter for analyzing the tendency of the y-axis recommendation total value for each of the x-axis item and the recommendation axis item is set,
The recommendation counting unit
From the calculated x-axis item and the calculated y-axis recommended value for each recommended axis item, using the statistical formula for quantifying the tendency set as the parameter in the recommendation method according to the recommendation method , Further calculating a recommendation evaluation value for each recommended axis item,
The data output unit is
The multidimensional data analysis support device which further outputs the recommendation evaluation value for each of the recommendation axis items calculated by the recommendation aggregation unit . The recommendation counting unit
The recommendation order of the recommendation axis item is calculated from the calculated magnitude relation of the recommendation evaluation value for each of the recommendation axis items,
The data output unit includes:
Wherein instead of the recommended evaluation value for each recommendation axis item, multidimensional data analysis support apparatus according to claim 1, further outputs the recommendation order of the recommended shaft items calculated by the recommendation summary section. Based on the x-axis item calculated by the recommended axis-specific intermediate totaling unit and the y-axis intermediate value for each recommended axis item, the value of the y-axis item for each x-axis item is calculated using the y-axis aggregation formula. And an integrated tabulation unit that calculates the y-axis integrated tabulation value.
The data output unit is
The multi-dimension according to claim 1 or 2 , further outputting a correlation between the x-axis item and the y-axis integrated total value from the y-axis integrated total value for each x-axis item calculated by the integrated total unit. Data analysis support device.

Images