Controls metadata
Metadata (Latin meta — the purpose, a terminal point, a limit, edge and data) — information on other information, or the data relating to the additional information about contents or an object. Metadata open data on the signs and properties characterizing any entities which allow to look for and manage automatically them in big information flows.
Content |
Databases
Such information is often used in databases:
- Metadata are subchannel information on the used data.
- The structured data representing characteristics of the described entities for the purposes of their identification, search, assessment, management of them. a set of the admissible structured descriptions which are available in an explicit form and which purpose can help to find an object. The term is used in the context of object search, entities, resources.
- Data from more general formal system, set from the system of data describing properties.
- Information on information which is contained on the web property page (creator, etc.). Example: A name of the author of editing in the text. This term is in the broadest sense used for any information on data: names of tables, columns in the table in relational databases, version number in the file of the program (i.e. as an informative part in the binary file) and t. p[1].
Difference between data and metadata
It is usually impossible to carry out unambiguous separation into data and metadata in the document as:
- Something can be both data, and metadata. So, the heading of article can be carried at the same time as to metadata (as an element of metadata — heading), and to actually this (as the heading is a part of the text).
- Data and metadata can change roles. On the poem considered as data music can be written, in this case all poem can be "attached" to the musical file and in this case be considered as metadata. Thus, reference to one or other category depends on the point of view (or namespaces, frames of reference).
- Creation meta-meta … - metadata is possible (see a choice axiom). As, according to normal determination, metadata are data, it is possible to create metadata on metadata, metadata for an output to special devices, or readings their description using the software transforming the text to the speech.
Other descriptive metadata can be used by the automated worker threads. For example, if some "smart" program "knows" contents and a data structure, then data can be automatically transformed and transferred to other "smart" program as input data. As a result, users will be exempted from need of accomplishment of a set of routine transactions if data are provided for work as such "laconic" programs.
Metadata become important in World Wide Web because of need of ensuring search of useful information among a huge number of available. The metadata created manually are of great value as it guarantees conciseness. If the web page on any certain subject contains a word or a phrase, then all other web pages on this subject may contain the same word or a phrase. Metadata also have a variety therefore if two values are connected with any subject, then each of them can be used. For example, article about LiveJournal can be designated using several values: "LiveJournal", LJ, LiveJournal.
Metadata are used for information storage about the records audio CD. Similar to MP3 files store metadata in ID3 format.
It is possible to edit metadata of graphic files in special programs for work with metadata.
Classification of metadata
Metadata can be classified on
- To contents. Metadata can or describe a resource (for example, the name and file size), or resource contents (for example, "in this video file it is shown as the guy plays soccer").
- In relation to a resource in general. Metadata can belong to a resource in general or to its parts. For example, "Title" (the title of the movie) belongs to the movie in general, and "Scene description" (the description of an episode of the movie) separate for each episode of the movie.
- Whenever possible logical output. Metadata can be subdivided into three layers: the low layer is "crude" data in itself; the center — the metadata describing the specified "crude" data; and a high layer — metadata which allow to draw a logical conclusion, using the second layer.
Three the most used classes of metadata are:
- The internal metadata describing structure or components of a thing, what the thing is. For example, format and file size.
- Administrative metadata which are required for processing of information, purpose of a thing. For example, information on the author, editor, date of publication, etc.
- Descriptive metadata which describe the thing nature, its signs. For example, a set of the categories connected with information, links to other things connected with this.
Management of metadata
Management of metadata - process of management of the data connected with data assets of the organization for ensuring integration, access, sharing, the analysis and so forth for achievement of the best results of the organization.
Controls metadata
- Metadata repositories (Repository of metadata) - is used for storage, documentation, the analysis and management of metadata. Comprises all data throughout all stages of project lifecycle: development, testing, input in commercial operation.
- Business glossary (Business the dictionary) - is used for management of business terms along with the corresponding determinations and the relations between these terms.
- Data lineage (Origin of data) - defines origin of data, stages of conversion and change of data, the direction of their movement.
- Impact analysis (The analysis of influence) is providing extensive information concerning dependences between data with the subsequent influence on target indicators.
- Rules management (Management of rules) is automation of observance of business rules which are tied to data members and the related metadata.
- Semantic frameworks (Semantic framework) is support of taxonomy, ER model, ontology, modeling of languages such as RDF, OWL and UML.
- Metadata ingestion and translation (Capture and transfer of metadata) is technologies and connectors for different data sources: RDBMS, BI, DIS, XML, etc.
Sources of metadata
Examples of standards of metamodels
Grouping of metadata
Practical application in Sberbank
Main objective of creation of uniform base of metadata is automation and quality improvement business protsessoviz [2]:
- Reduction in cost of the analysis and design of solutions
- Reduction of time of development and output of the code on Wednesdays
- Quality improvement of products
- Control of compliance to architectural requirements
The uniform base of metadata, integrates analysis results and architecture of project solutions with metadata of real environments.
The analytical space, is the high-level description of a path of data from the systems of sources to show-windows. It is developed at an analysis stage and design of the solution.
The directory information a component, provides the user with the complete, operational and systematized information about used in the TsHD complex – objects, communications between objects, components, metrics, placement.
Notes
- ↑ Material from Wikipedia — the free encyclopedia
- ↑ Selivanov Yu.A. presentations . - Sberbank-Technologies (SberTech) on [1] TAdviser SummIT 2017