Multi-level data structuring and analysis system

Abstract

The present invention discloses a multi-level data structuring and analysis system comprising a data upload module (110) for uploading raw data including multi-level hierarchical variables, a data conversion module (120) for reading raw data and converting the read data into a data structure capable of statistical analysis, and a report generation module (130) for analyzing the converted data according to predefined analysis items and generating a structured result report, wherein the uploading, conversion, analysis, and result report generation of data are carried out as a one-stop process through a built-in specific (predetermined) algorithm.

Description

Technology Field

[0001] The present invention relates to a multi-level data structuring and analysis system, and more specifically, to a multi-level data structuring and analysis system that enables data uploading, conversion, analysis, and result report generation to proceed as a one-stop process through a built-in specific (predetermined) algorithm. Background Technology

[0002] As is well known, the analysis results are organized into multiple result tables, and the format of the result tables is specified; the analysis mostly consists of descriptive statistics and does not require complex inferential statistics.

[0003] However, there are limitations to creating report result tables using only Excel pivots, and there are many result tables that absolutely require manual work.

[0004] In addition, even after creating the analysis results table, it is necessary to manually edit the row and column sizes, fonts, etc., of all tables.

[0005] For example, the Korean Hospital Pharmacists Association outsources data analysis and report writing every year, but the same research team from the previous year may or may not win the contract, and it is difficult to accurately identify the requirements for some result table items due to the report format and footnotes.

[0006] As such, conventional data analysis tools such as Microsoft Excel have limitations in manual preprocessing and variable selection. When raw data consists of multi-level hierarchical variables or is in wide form, visual data exploration is difficult, and it is not possible to selectively select only a specific sub-level of a variable. Although the pivot function provided by Excel is very useful, to use it with data having the aforementioned structural characteristics, a preprocessing step (data handling step) is required to convert it into a single-level structure variable to which the pivot function can be applied. Furthermore, in the case of datasets with a large number of variables, generating variable names is also burdensome. This data preprocessing process occurs commonly when using not only Excel but also other general-purpose analysis packages such as Python, R, SAS, SPSS, and STATA.

[0007] In addition, operational definitions or interpretations required for analysis may vary from researcher to researcher, and relying on arbitrary judgment or past performance memory consumes unnecessary time to understand and verify them. Furthermore, since each researcher generates data variable names differently, comparative analysis between data is not smooth. Moreover, in order to document analysis results, the output from Excel or statistical packages is transferred to a word processor program, and the editing process to improve readability is performed manually, resulting in inefficiency of the entire process.

[0008] In addition, when presenting statistical data by multiple factors in a single results table, using Excel or other statistical software presents the inconvenience of having to perform analyses by factor (variable) and then merge the results, or even if the results can be obtained at once, manually edit them to fit the report format. Prior art literature

[0009] Korean Published Patent Application No. 10-2025-0017834 (Device and method for generating a data analysis report using a data storytelling method and a recording medium storing a program for implementing the same, Feb. 4, 2025) Korean Registered Patent Application No. 10-2022944 (Method for automatically generating a big data analysis report and device for performing the same, Announced Sep. 20, 2019) The problem to be solved

[0010] The technical problem that the concept of the present invention aims to solve is to provide a multi-level data structuring and analysis system capable of ensuring the reliability and accuracy of data processing and analysis by consistently deriving analysis results through a one-stop process of uploading, transforming, analyzing, and generating result reports of large-scale data composed of multi-level hierarchical variables using a specific (predetermined) algorithm. means of solving the problem

[0011] To achieve the aforementioned objective, one embodiment of the present invention provides a multi-level data structuring and analysis system comprising: a data upload module for uploading raw data including multi-level hierarchical variables; a data conversion module for reading the raw data and converting the read data into a data structure capable of statistical analysis; and a report generation module for analyzing the converted data according to predefined analysis items and generating a structured result report, wherein the uploading, conversion, analysis, and result report generation of the data are performed as a one-stop process through a built-in algorithm.

[0012] Here, the data conversion module can convert the read data by reconstructing it into a single-layer structure.

[0013] In addition, the data conversion module can select variables and individual data constituting each variable according to analysis requirements and convert them into a sub-dataset form that allows the use of Excel pivots.

[0014] In addition, the above analysis requirements may include an operational definition of each variable and a predetermined calculation formula for calculating the analysis items of the said variable according to the said operational definition.

[0015] In addition, the above algorithm includes a pre-specified document format and can generate the result report according to the specified document format.

[0016] In addition, the above variables may be selected in detail by hierarchical level.

[0017] In addition, the data upload module uploads the raw data in wide-form format, and the data conversion module can convert the raw data by reconstructing the hierarchical depth of the variables into a single-layer structure.

[0018] In addition, the data conversion module can provide a UI by visualizing the data structure as a tree structure.

[0019] In addition, the above raw data may be data from a survey of the actual conditions of hospital pharmacy departments. Effects of the invention

[0020] According to the present invention, through a specific (predetermined) algorithm, the uploading, conversion, analysis, and generation of result reports of large-scale data consisting of multi-level hierarchical variables are carried out as a one-stop process, thereby drastically reducing the time required for analysis and report writing, enabling the generation of complex summary tables with a single operation, enabling rapid response even if a situation involving the updating of raw data including the modification of individual data occurs after the start of analysis, enabling rapid comparison with data before and after updates, before and after changes, or previous years, and minimizing document editing.

[0021] In addition, it has the effect of ensuring the reliability and accuracy of data processing and analysis by consistently deriving the analysis results required by KSHP (Korean Society of Hospital Pharmacists) without relying on the memory and judgment of the person in charge or the researcher.

[0022] Furthermore, beyond the purpose of preparing annual reports, this system supports convenient data processing, such as variable selection and sub-dataset creation as needed. This facilitates various analyses utilizing data containing multi-level hierarchical variables, reduces the cost of outsourcing result analysis and report preparation each year, and provides a one-stop process UI screen to integrate data editing, analysis, and result report generation. It also supports users in intuitively editing and managing data, thereby minimizing the data preprocessing process. Additionally, all files loaded into the system are listed for selection, and if necessary, it is possible to replace the currently used data file with another for analysis and compare results with existing data without closing the file. Brief explanation of the drawing

[0023] Figure 1 illustrates a configuration diagram of a multi-level data structuring and analysis system according to one embodiment of the present invention. Figure 2 illustrates the raw data of the multi-level data structuring and analysis system of Figure 1. Figure 3 illustrates data uploading through the UI screen of a process for implementing the multi-level data structuring and analysis system of Figure 1. Figure 4 illustrates the creation of a sub-dataset in the UI screen of Figure 3. Figure 5 illustrates a single-layer structure transformation of data in the UI screen of Figure 3. Figure 6 illustrates the generation of a result report on the UI screen of Figure 3. FIG. 7 illustrates a flowchart of a multi-level data structuring and analysis method according to another embodiment of the present invention. Specific details for implementing the invention

[0024] Hereinafter, embodiments of the present invention having the aforementioned features will be described in more detail with reference to the attached drawings.

[0025] A multi-level data structuring and analysis system according to one embodiment of the present invention comprises a data upload module (110) for uploading raw data including multi-level hierarchical variables, a data conversion module (120) for reading raw data and converting the read data into a data structure capable of statistical analysis, and a report generation module (130) for analyzing the converted data according to predefined analysis items and generating a structured result report, wherein the uploading, conversion, analysis, and result report generation of data are carried out as a one-stop process through a built-in specific (predetermined) algorithm.

[0026] Hereinafter, with reference to the drawings, the multi-level data structuring and analysis system of the above-described configuration will be specifically described as follows.

[0027] First, the data upload module (110), referring to FIGS. 1 and FIGS. 2, uploads raw data containing multi-level hierarchical variables to the DB.

[0028] Here, the raw data may be data from the status survey of hospital pharmacy departments conducted annually with the same items or from the abbreviated survey items in odd-numbered years, but is not specifically limited to this and may be various data including multi-level hierarchical variables.

[0029] For example, the data upload module (110) can upload raw data of wide-form data consisting of 5 levels of hierarchical variables collected through the process of querying and verifying the survey questions of the questionnaire for the hospital pharmacy department survey, as shown in FIG. 2.

[0030] In addition, referring to Fig. 3, on the UI screen of the one-stop process for uploading, converting, analyzing, and generating result reports on data, you can click the 'Upload Data' menu to select and upload a raw data file.

[0031] In addition, when uploading data, it is possible to specify the data filename on the UI screen and add notes regarding data characteristics or issues.

[0032] Next, the data conversion module (120), referring to FIG. 1, reads the raw data, converts the read data into a data structure capable of statistical analysis, and selects columns to perform filtering operations for generating a result report.

[0033] Specifically, the data conversion module (120) can convert the read data into a single-layer structure by clicking the 'Edit Data' menu on the UI screen. The data conversion module (120) can generate a sub-data set that can be exported as an Excel file and used immediately as an Excel pivot, by selecting variables and individual data constituting each variable according to the analysis requirements for the analysis of the raw data variables of Figure 2 as shown in Figure 4 (a), as shown in Figure 5 (b).

[0034] Here, the data conversion module (120) can select and reconstruct only the necessary variables / data, and can also select only specific lower levels of multi-level hierarchical variables, so that when sharing data, sensitive information can be excluded or only necessary variables can be selected to create a file, and various exploratory analyses or trend verifications can be performed.

[0035] Referring to Figure 4 (b), the data conversion module (120) can provide a UI by visualizing the data structure as a tree structure that preserves the hierarchical attribute information of the original data.

[0036] In addition, the aforementioned analysis requirements may include the operational definition of each variable and a specific (predetermined) calculation formula for calculating the analysis items of the variable according to the operational definition.

[0037] For example, an operational definition refers to the process of quantifying data so that it can be objectively measured or transforming it into clear standards. By applying operational definitions to each variable, such as hospital staff, the number of patients, and drug usage, arbitrary definitions by the analysis subject, such as researchers, are excluded, ensuring that the same analysis results are derived even if the analysis subject changes. Furthermore, even if sub-datasets change year by year, the consistency and reliability of analysis results can be guaranteed based on pre-defined operational definitions using the same standards, and manual errors can be minimized.

[0038] In addition, as shown in FIG. 5 (a), variables can be selected in detail by the user at the hierarchical level, and as shown in FIG. 5 (b), the data conversion module (120) can convert the hierarchical depth of the variables of the original data into a single-layer structure and export it as an Excel file.

[0039] Next, the report generation module (130), referring to FIG. 1, analyzes the converted data according to predefined analysis items and generates a structured and standardized result report.

[0040] For example, as shown in FIG. 6 (a), the report generation module (130) lists result tables through a UI screen, and by selecting some or all of the result tables, as shown in FIG. 6 (b), a structured result report without human error can be generated without relying on the manual method of the analysis subject.

[0041] In addition, a specific (specified) algorithm includes a pre-specified document format, and can generate a result report according to the specific (specified) document format, making it usable as a deliverable that can be distributed immediately upon implementation.

[0042] FIG. 7 illustrates a flowchart of a multi-level data structuring and analysis method according to another embodiment of the present invention. The multi-level data structuring and analysis method according to another embodiment of the present invention includes a data upload step (S110) for uploading raw data including multi-level hierarchical variables through a data upload module (110), a data conversion step (S120) for reading raw data and converting the read data into a statistically analyzable data structure through a data conversion module (120), and a report generation step (S130) for generating a structured result report by analyzing the converted data according to a predefined analysis item through a report generation module (130), thereby enabling the uploading, conversion, analysis, and result report generation of data to proceed as a one-stop process through a built-in specific (predetermined) algorithm.

[0043] Specifically, prior to this, in the data upload step (S110), the data upload module (110) uploads raw data containing multi-level hierarchical variables to the DB, with reference to FIGS. 1 and FIGS. 2.

[0044] Here, the raw data may be data from the status survey of hospital pharmacy departments conducted annually with the same items or from the abbreviated survey items in odd-numbered years, but is not specifically limited to this and may be various data including multi-level hierarchical variables.

[0045] For example, the data upload module (110) can upload raw data of wide-form data consisting of 5 levels of hierarchical variables collected through the process of querying and verifying the survey questions of the questionnaire for the hospital pharmacy department survey, as shown in FIG. 2.

[0046] In addition, referring to Fig. 3, on the UI screen of the one-stop process for uploading, converting, analyzing, and generating result reports on data, you can click the 'Upload Data' menu to select and upload a raw data file.

[0047] In addition, when uploading data, it is possible to specify the data filename on the UI screen and add notes regarding data characteristics or issues.

[0048] Subsequently, in the data conversion step (S120), the data conversion module (120), referring to FIG. 1, reads the raw data, converts the read data into a data structure capable of statistical analysis, and may select columns to perform filtering operations for generating a result report.

[0049] Specifically, the data conversion module (120) can convert the read data into a single-layer structure by clicking the 'Edit Data' menu on the UI screen. The data conversion module (120) can generate a sub-data set that can be exported as an Excel file and used immediately as an Excel pivot, by selecting variables and individual data constituting each variable according to the analysis requirements for the analysis of the raw data variables of Figure 2 as shown in Figure 4 (a), as shown in Figure 5 (b).

[0050] Here, the data conversion module (120) can select and reconstruct only the necessary variables / data, and can also select only specific lower levels of multi-level hierarchical variables, so that when sharing data, sensitive information can be excluded or only necessary variables can be selected to create a file, and various exploratory analyses or trend verifications can be performed.

[0051] Referring to Figure 4 (b), the data conversion module (120) can provide a UI by visualizing the data structure as a tree structure that preserves the hierarchical attribute information of the original data.

[0052] In addition, the aforementioned analysis requirements may include the operational definition of each variable and a specific (predetermined) calculation formula for calculating the analysis items of the variable according to the operational definition.

[0053] For example, an operational definition refers to the process of quantifying data so that it can be objectively measured or transforming it into clear standards. By applying operational definitions to each variable, such as hospital staff, the number of patients, and drug usage, arbitrary definitions by the analysis subject, such as researchers, are excluded, ensuring that the same analysis results are derived even if the analysis subject changes. Furthermore, even if sub-datasets change year by year, the consistency and reliability of analysis results can be guaranteed based on pre-defined operational definitions using the same standards, and manual errors can be minimized.

[0054] In addition, as shown in FIG. 5 (a), variables can be selected in detail by the user at the hierarchical level, and as shown in FIG. 5 (b), the data conversion module (120) can convert the hierarchical depth of the variables of the original data into a single-layer structure and export it as an Excel file.

[0055] Subsequently, in the report generation step (S130), the report generation module (130), referring to FIG. 1, analyzes the converted data according to predefined analysis items and generates a structured and standardized result report.

[0056] For example, as shown in FIG. 6 (a), the report generation module (130) lists result tables through a UI screen, and by selecting some or all of the result tables, as shown in FIG. 6 (b), a structured result report without human error can be generated without relying on the manual method of the analysis subject.

[0057] In addition, a specific (specified) algorithm includes a pre-specified document format, and can generate a result report according to the specific (specified) document format, making it usable as a deliverable that can be distributed immediately upon implementation.

[0058] Meanwhile, another embodiment of the present invention provides a computer-readable recording medium having a program recorded thereon for executing the above-described multi-level data structuring and analysis method on a computer.

[0059] Accordingly, by utilizing a specific algorithm to carry out the uploading, transformation, analysis, and generation of result reports for large-scale data composed of multi-level hierarchical variables as a one-stop process, the time required for analysis and report writing is drastically reduced; complex summary tables can be generated with a single operation; rapid responses are possible even if raw data updates, including modifications to individual data, occur after the analysis begins; quick comparisons are possible with data before and after updates, before and after changes, or with previous years; document editing is minimized; and reliability and accuracy of data processing and analysis can be ensured by consistently deriving analysis results required by KSHP (Korean Society of Hospital Pharmacists) without relying on the memory and judgment of the person in charge or the researcher. Furthermore, beyond the purpose of preparing annual reports, support is provided for convenient data processing such as variable selection and sub-dataset creation as needed, thereby facilitating various analyses utilizing data containing multi-level hierarchical variables and reducing external outsourcing costs for result analysis and report writing each year. Additionally, by providing a UI screen for the one-stop process, data editing, analysis, and result report generation can be performed in an integrated manner. It is possible to minimize the data preprocessing process by supporting users to intuitively edit and manage data, and all files loaded into this system are listed and selectable. If necessary, it is possible to replace the data file currently in use with another data file for analysis and compare it with existing data results without closing the current data file.

[0060] The embodiments described in this specification and the configurations illustrated in the drawings are merely the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention; therefore, it should be understood that various equivalents and modifications that can replace them may exist at the time of filing this application. Explanation of the symbols

[0061] 110: Data Upload Module 120 : Data conversion module 130: Report Generation Module

Claims

Claim 1 A multi-level data structuring and analysis system comprising: a data upload module for uploading raw data including multi-level hierarchical variables; a data conversion module for reading the raw data and converting the read data into a data structure capable of statistical analysis; and a report generation module for analyzing the converted data according to predefined analysis items and generating a structured result report; wherein the uploading, conversion, analysis, and result report generation of the data are carried out as a one-stop process through a built-in algorithm. Claim 2 A multi-level data structuring and analysis system according to claim 1, wherein the data conversion module reconstructs and converts the read data into a single-layer structure. Claim 3 A multi-level data structuring and analysis system according to claim 1, wherein the data conversion module selects variables and individual data constituting each variable according to analysis requirements and converts them into a sub-dataset form capable of using Excel pivot. Claim 4 A multi-level data structuring and analysis system according to claim 3, wherein the analysis requirements include an operational definition of each of the above variables and a predetermined calculation formula for calculating an analysis item of the said variable according to the said operational definition. Claim 5 A multi-level data structuring and analysis system according to claim 1, wherein the algorithm includes a pre-specified document format and generates the result report according to the predetermined document format. Claim 6 A multi-level data structuring and analysis system characterized in that, in claim 3, the above variables can be selected in detail by hierarchical level. Claim 7 A multi-level data structuring and analysis system according to claim 1, wherein the data upload module uploads the raw data of wide-form data, and the data conversion module converts the variable hierarchy depth of the raw data into a single-layer structure. Claim 8 A multi-level data structuring and analysis system according to claim 7, wherein the data conversion module provides a UI by visualizing the data structure as a tree structure. Claim 9 A multi-level data structuring and analysis system according to claim 1, characterized in that the raw data is actual status survey data of a hospital pharmacy department.

Images