In simple but technical term, metadata is a data that describes another data. It can be any item describing an individual datum or a collection of multiple content items.
Metadata is very useful in facilitating the use, management and understanding of data in a large data warehouse. Depending on the type of data and the context where the data is being used, metadata required to effectively manage a database or large data warehouse varies.
For instance in a library system, metadata to be used will surely include description of book contents, authors, data of publication and the physical location of books in the library. If the context of use is about photography, the metadata which will be used are for description of the camera, model, types, photographer, date photograph was taken, location where photograph is taken and many other things. In the case of an information system where data involved is the content of files in the computer, metadata used will be describing individual data items and their field names, length, etc.
Common Business Metadata is one of the foundations of an intelligent business system. Common Data Names, Common Data Definitions and Common Data Integrity rules need to be very consistent. Trustworthiness and common understanding are important prerequisites for integrating business intelligence into operational business processes. Without those perquisites, integrating business intelligence into business operations can potentially make untold damage to a high degree.
A Common Metadata is the bases for sharing data within an enterprise. These data refer to a common definition of data items, Common Data Names and Common Integrity Rules. With these commonalities come Common Transformations for all master data items including customer, employee, product, location and many others. This also includes Common Transformations for all business transaction data and all business intelligence system metrics.
In a data warehouse, it is extremely important to have a Common Warehouse Metamodel. This model specifies the modeling aspect of metadata which are being used for relational, non-relational, multi-dimensional and other objects found within the data warehouse so that the system will have a Common Metadata Structure that adheres to the underling business data architecture.
Common interfaces that can be useful in enabling interchange of business and warehouse intelligence metadata are specified within the Common Warehouse Metamodel. This can be in conjunction with warehouse tools, warehouse metadata repositories and data warehouse platforms in heterogeneous distributed warehouse environments. A Common Warehouse Metamodel is based on three standard which are Unified Modeling Language (UML), Meta Object Facility (MOF) and XML Metadata Interchange (XMI).
Common Warehouse Metamodels are also useful in enabling users to trace data lineage as they provide objects that effectively describe where the data came from and how or when the data is being created. The instance of the metamodel are exchanged through the XML Meta Data Interchange documents.
Today’s business trends are heading towards the internet as the main highway to gather and share data. But the internet is full of all sorts of data. This includes different data formats, different applications using and sharing data and different server systems. Problems can arise in terms of hardware and software portability.
The use of Common Metadata tries to melt the boundary down because the format by which a common data is packaged can be read by disparate systems. So, whether the data shared in used a relational database or an excel flat files, the processing server within the data warehouse will know how to deal with data for processing.