CN107609155B - Construction method of data asset model based on XBRL standard - Google Patents

Abstract

The invention belongs to the technical field of computer financial standardization software, and discloses a method for constructing a data capitalization model based on XBRL (extensible business reporting language) standard, which solves the problems of disordered data architecture and data standard loss which are often encountered in the process of performing data capitalization management in a big data era by a traditional enterprise. The method comprises the following steps: a. carrying out element information preprocessing and labeling on the original business bill; b. marking the preprocessed marked bills in further detail to generate an MDS file; c. generating an XBRL intermediate format file XDS which is convenient for a computer to read by the MDS through a converter; d. generating classification standards in XBRL format and example documents by the XDS through a converter; e. and storing the classification standard and the example document in the XBRL format in a relational database. The method is suitable for constructing the data asset model of the XBRL standard.

Description

Construction method of data asset model based on XBRL standard

Technical Field

The invention belongs to the technical field of computer financial standardization software, and particularly relates to a construction method of a data capitalization model based on an XBRL standard.

Background

Data capitalization is a significant proposition for improving data value in the current big data era. According to the research results of international authorities, the data asset management generally comprises the following steps: firstly, data management; constructing a data structure; thirdly, developing data; the data is safe; main data; sixthly, a data warehouse; data quality; data standard; metadata, etc.

Most of the traditional enterprises have various problems in the process of data asset management in the era of big data transformation. For example, data structure confusion: the number of systems is increased, the complexity of the systems is increased, the management difficulty is increased, and no one can know the data architecture and the data flow direction of the whole system. Standard deletion of data: the data model meaning is difficult to understand for development and operation and maintenance personnel due to the lack of uniform data standards at the enterprise level, so that data integration and data sharing among different applications of an enterprise are difficult.

Disclosure of Invention

The technical problem to be solved by the invention is as follows: a method for constructing a data capitalization model based on an XBRL standard is provided, and the problems of disordered data architecture and data standard loss which are often encountered in the data capitalization management process of a traditional enterprise in a big data era are solved.

The technical scheme adopted by the invention for solving the technical problems is as follows:

a construction method of a data asset modeling model based on XBRL standard comprises the following steps:

a. carrying out element information preprocessing and labeling on the original business bill;

b. marking the preprocessed marked bills in further detail to generate an MDS file;

c. generating an XBRL intermediate format file XDS which is convenient for a computer to read by the MDS through a converter;

d. generating classification standards in XBRL format and example documents by the XDS through a converter;

e. and storing the classification standard and the example document in the XBRL format in a relational database.

As a further optimization, in the step a, when the element marking is performed, the Property file is used as a reference, and the marking process is to mark the element in the bill with the row number of the corresponding element in the Property file; the Property file is a file in an Excel format, wherein the description of field elements and document use elements of an original document is provided, and the Property file comprises some most basic element attributes: the name, data type, length, and business module of the element.

As a further optimization, in step b, the step of further detailed labeling of the pre-processed labeled bill includes:

firstly, element structure labeling is carried out on a single bill by adopting Mini-MDS, and then the labeled Mini-MDS is combined into Master-MDS.

As a further optimization, the Mini-MDS performs element structure labeling on a single bill, including definition of a document element semantic part, the content of which includes: 1) the document model goes from an abstract layer to a concrete layer; 2) a document tree structure; 3) Chinese-English interpretation of document nodes; 4) the corresponding relationship between the document element node and the original bill element; 5) length and type of document element.

For further optimization, the Master-MDS is used for uniformly containing all the defined Mini-MDSs in one file, and simultaneously contains all meta-data for generating XBRL classification standards.

As a further optimization, in step c, the merged MDS is converted into an XBRL intermediate format document XDS by a converter Transformer M-X, where the XDS contains the necessary content for creating XBRL classification criteria.

As a further optimization, in step d, the XDS is converted into document model classification standard DMT in XBRL format by converter Transformer X-T, i.e. XBRL classification standard;

the pattern files contained in the generated DMT are:

1) a part pattern file; 2) a role mode file; 3) a type schema file; 4) a core schema file;

the linkbase files contained in the generated DMT are:

1) defining a link library; 2) a tag link library; 3) tag enumeration link library; 4) listing a link library; 5) referring to the defined attribute link library; 6) referencing an element mapping link library; 7) referring to a sample fact value link library; 8) the linked library is referenced with reference to the extension definition.

As a further optimization, in step d, DMT can be converted into an instance document in XBRL format by concatenating a JSON formatted data source from an XDS file via a transform X-S.

As a further optimization, the transformers M-X, transformers X-T and ransformer X-S are script programs used at different stages.

The invention has the beneficial effects that: by adopting the XBRL technical standard, all documents, bills and vouchers of different types of enterprises can be converted into corresponding XBRL classification standards and XBRL instance documents, and the method has the effects of unification of enterprise data standards and clear data flow processing architecture.

Detailed Description

The invention aims to provide a construction method of a data capitalization model based on XBRL standard, which solves the problems of disordered data architecture and data standard loss which are often encountered in the data capitalization management process of the traditional enterprise in the big data era.

The invention adopts the internationally recognized financial information data exchange standard and technology-XBRL, and forms a model Definition file MDS (model Definition Spreadsheet) by analyzing and manually modeling data on a document from a service original document object. The MDS obtains an XBRL Definition document XDS (XBRL Definition Spreadsheet) through first automatic conversion. Then, the classification standard in the XBRL format is obtained through automatic conversion, and an example document template is generated at the same time. And finally, completing the modeling process from the original business data to the data asset standard.

In particular implementation, the method for constructing the data asset modeling model based on the XBRL standard comprises the following steps:

firstly, preprocessing and labeling an original document:

the original document can be a paper bill or a scanned image of the bill, and is the most direct and intuitive business scene.

The method is characterized in that element labels of an original document are established, how the elements are combed and what elements are combed are established, and the basic information is firmly held in the hands of the enterprise instead of relying on external software.

And marking each element needing to be modeled on the document. The file referred to by the label is called a Property file, which is a file in an Excel format, and the description of elements such as field elements, document purposes and the like of an original document is arranged in the Property file. Some of the most basic element attributes are included, such as the name of the element, the data type, the length, and the business module to which it belongs. Provided by the service party card.

The marking is to mark the element in the bill with the line number of the corresponding element in the Property file. The method has the advantage that whether the basic business elements are accurately marked or not or are not marked can be conveniently searched. This is the most fundamental work that can be done by service personnel under the direction of a technician. Professional XBRL marks can be added, and elements of the sub-businesses are assembled into an element library.

Secondly, detailed labeling processing based on the MDS file:

the model Definition file MDS (model Definition Spreadsheet) adopts an Excel format file, and each element of the original document is subjected to more detailed professional annotation explanation on the basis of the Property file, so that support is provided for subsequent processing. The detailed labels include the elementary label semantics of the elements, the model structure type, the module hierarchy, the module sequence number, and the like.

For collaboration among team members, we split MDS into Mini-MDS and Master-MDS.

The Mini-MDS is used for carrying out model annotation on a single different bill, mainly comprises the definition of a semantic part of a document element, and the content mainly comprises the following components: 1) the document model goes from an abstract layer to a concrete layer; 2) a document tree structure; 3) Chinese-English interpretation of document nodes; 4) the corresponding relationship between the document element node and the original bill element; 5) length and type of document element. For example: the train ticket can be marked as a Mini-MDS document; the general quota invoice may be marked as another Mini-MDS document. Therefore, the carding marking work of the original bill with large workload can be distributed to a plurality of people to be carried out simultaneously.

If a Mini-MDS annotation is completed, it can be copied to Master-MDS. The Master-MDS has only one copy, and preferably, one person can uniformly maintain and manage the Master-MDS. After the Master-MDS is completed, it can then be converted into a version of the computer recognition process.

Master-MDS has multiple sheet pages. The main sheet pages are: ModelMaster, PackageOrders, PackageIndex, XBRL DTS, XBRL Columns, EnumetetdTypesIndex, EnumelationsList.

Mini-MDS is copied to the sheet page ModelMaster of Master-MDS. The tail of the head list of the ModelMaster page is numbered in sequence with PkgA-Order, PkgA-Depth, PkgB-Order and PkgB-Depth, and respectively represents different bill documents. If an element row appears in a ticket, the element's serial number, hierarchical depth, is noted clearly under the corresponding Pkg column.

The method comprises the steps of service marking of an original document, element structure marking of the Mini-MDS and detailed marking of XBRL information of the Master-MDS, and therefore, the method is easy to Master by an enterprise user in a modeling mode of one step from shallow to deep, progressive and ring-and-ring buckling, and each step is visible and touchable and is easy to modify and perfect.

Thirdly, generating an XDS (extensible Business support service) of the XBRL intermediate format file:

XBRL Definition document XDS (XBRL Definition Spreadsheet) is transformed from MDS by using a first Transformer M-X. XDS is also in EXCEL format. The XDS and MDS descriptions are the same, except that MDS is directly established manually and is suitable for being understood and read by people; the XDS is generated by a converter and contains some information suitable for computer reading required to ultimately generate the XBRL taxonomy.

The XDS contains the necessary content to create the XBRL classification criteria. The XDS file includes XBRL Concepts pages, as well as XBRL DTS pages.

The conversion from MDS to XDS is just an intermediate link in the conversion process of the script program, and is convenient for professional technicians to learn and use by mistake. The average business person can skip directly, keeping this intermediate result, and it can be seen that it is an especially elegant elaboration of the modeling process.

Fourthly, generation of XBRL classification standards:

from the XDS document, the XDS can be converted to the document Model Classification Standard DMT (document Model Taxomy), also known as XBRL Classification Standard, in XBRL format by a second converter, transducer X-T.

The XBRL classification standard is the latest technical standard which is recognized in the industry and realizes data integration based on internet and cross-platform operation. Once recording and multiple use. Can be used for automatic data exchange without the limitation of individual company software and information systems, and provides convenience for investors or analysts to use financial information. Thus, the conversion from XDS to XBRL classification criteria is of great benefit. It is not described in detail.

The number in the DMT file name is the version number of the associated master and the date of its release.

The pattern files contained in the generated DMT are:

1) a part pattern file:

changhong-XDC_parts_v0.1.3_2017-07-26.xsd ；

2) role mode file:

changhong-XDC_roles_v0.1.3_2017-07-26.xsd ；

3) type mode file:

changhong-XDC_types_v0.1.29_2017-07-26.xsd ；

4) a core mode file:

XDC_core_v0.1.29.xsd 。

the linkbase files contained in the generated DMT are:

1) defining a link library:

XDC_core_v0.1.29_2017-07-26_def.xml ；

2) and (3) tag link library:

XDC_core_v0.1.29_2017-07-26_lab.xml ；

3) tag enumeration link library:

XDC_core_v0.1.29_2017-07-26_lab_enum.xml ；

4) listing a link library:

XDC_core_v0.1.29_2017-07-26_pre.xml ；

5) referring to the defined attribute link library:

XDC_core_v0.1.29_2017-07-26_ref_definitionalAttribute.xml ；

6) referring to the element mapping link library:

XDC_core_v0.1.29_2017-07-26_ref_elementMapping.xml ；

7) referring to the sample fact value link library:

XDC_core_v0.1.29_2017-07-26_ref_exampleFact.xml ；

8) referencing the extended definition to the linked library:

XDC_core_v0.1.29_2017-07-26_ref_externalDefinitionReference.xml。

fifthly, generating the XBRL instance document:

light has XBRL classification criteria, no data, and is also a flower shelf. Therefore, after the classification criteria are generated in the previous step, DMT can be converted into an instance document in XBRL format by connecting the data sources (JSON format) from the XDS file through a third converter, transducer X-S, which is used for conversion into a relational database and stored in the XDC. The entire modeling process ends.

The XBRL instance document has the advantages of the XBRL system architecture and is very beneficial. The XBRL instance document can be conveniently converted into written text, a PDF file, an HTML page or other corresponding file formats. The information obtained by the XBRL can be conveniently applied to various fields of financial analysis and the like without printing or inputting again. The XBRL international organization is added to the four registered accounting firms, fuji, microsoft, IBM, SAP and other famous companies all around the world, and the continuous development of the technology is promoted.

Because a practical and efficient XML database does not exist at present, the XBRL instance document needs to be converted into a relational database again to be stored, and the relational database is called as XDC in the system. Although it is also stored in the relational database, it has a significant difference from the general relational database, that is, the XDC is with XBRL property and can be easily restored to XBRL instance document, XBRL classification standard. Therefore, the XBRL characteristic of the modeling is not reduced.

In the processes of generating the XBRL intermediate format document XDS document, the XBRL classification standard and the XBRL instance document, a converter is involved, and different conversion actions are completed in different stages inside the program.

The use method of the converter comprises the following steps:

1) download and install the Arelle tool. (if the installation position is G: \ Arelle)

2) Copy the transformation module files model extract, loadfromcel, savampleinstant, py into the plugin subdirectory under the Arelle's installation directory.

3) And editing a script file runGen _ XDC-XBRL.

4) Running a script program, and carrying out correct parameters: MDS version number, roles file version number, parts file version number.

5) Generating an output folder: XDC _ XBRL _ v + MDS version number. The folder contains sorting criteria and instance documents.

Claims (4)

1. A construction method of a data asset modeling model based on XBRL standard is characterized by comprising the following steps:

a. carrying out element information preprocessing and labeling on the original business bill;

b. marking the preprocessed marked bills in further detail to generate a model definition file MDS, wherein the method comprises the following steps:

firstly, element structure labeling is carried out on a single bill by adopting Mini-MDS, and then the labeled Mini-MDS is merged into Master-MDS, wherein the Master-MDS is used for uniformly containing all the defined Mini-MDS in one file, and simultaneously the Master-MDS contains all meta-data for generating XBRL classification standards;

c. generating an XBRL intermediate format file XDS which is convenient for a computer to read by an MDS file through a converter Transformer M-X;

d. generating an XDS file into a classification standard in an XBRL format and an example document through a converter Transformer X-T;

e. storing the classification standard and the example document in the XBRL format in a relational database;

in the step c, the merged MDS file is converted into an XBRL intermediate format file XDS through a converter Transformer M-X, wherein the XDS file contains necessary contents for creating XBRL classification standards;

in the step d, converting XDS into a document model classification standard DMT in an XBRL format through a converter Transformer X-T, namely the XBRL classification standard; connecting a data source in a JSON format from an XDS file through a Transformer X-S, and converting DMT into an example document in an XBRL format;

the Mini-MDS carries out element structure labeling on a single bill, and the element structure labeling comprises definition of a document element semantic part, and the content of the element structure labeling comprises the following steps: 1) the document model goes from an abstract layer to a concrete layer; 2) a document tree structure; 3) Chinese-English interpretation of document nodes; 4) the corresponding relationship between the document element node and the original bill element; 5) length and type of document element.

2. The method according to claim 1, wherein in the step a, when the element marking is performed, the Property file is used as a reference, and the marking process is to mark the element in the bill with the row number of the corresponding element in the Property file; the Property file is a file in an Excel format, wherein the description of field elements and document use elements of an original document is provided, and the Property file comprises some most basic element attributes: the name, data type, length, and business module of the element.

3. The method of claim 1, wherein the pattern file included in the DMT generated in step d comprises:

1) a part pattern file; 2) a role mode file; 3) a type schema file; 4) a core schema file;

the linkbase files contained in the generated DMT are:

1) defining a link library; 2) a tag link library; 3) tag enumeration link library; 4) listing a link library; 5) referring to the defined attribute link library; 6) referencing an element mapping link library; 7) referring to a sample fact value link library; 8) the linked library is referenced with reference to the extension definition.

4. The method of claim 1, wherein the transformers Transformer M-X, Transformer X-T and ransformer X-S are scripting programs used at different stages.