Data Structure represents both physical and logical data contained in common data architecture.  It includes data entities, data subjects, its contents, relationships and arrangement.

Data component refers to a component of the metadata warehouse that contains the structure of data within the common data architecture.

In general, 

• data structures define what and how data will be stored in the database or data warehouse 
• how long the data will be stored 
• if it is no longer needed how will it be disposed of or archived 
• who will be responsible for collecting and ensuring quality 
• who will has access to data

A real enterprise wide data warehouse has very complex data architecture. A data warehouse is a repository of all enterprise related data coming from various data sources such are those coming from different departments (i.e. Finance, Administrative, Human Resource Departments).

Some of the high volumes of data will be stored in large legacy of package system wherein data structure may be unknown. Other enterprise data may be contained in spreadsheets and smaller personal databases such as Microsoft’s Access and these aforementioned data may not be known by the IT department.

Some of the key information may be residing in some external information systems which are maintained by third party service providers or business partners.

Without a well defined data architecture and data structure, there can be very little control over the realization of high level business data concepts while data will likely be highly dispersed and of poor quality.

Another negative effect is that most data be redundant across the system and may result in conflicts in the organizational and business processes.

Data structures can be defined with high level data models which describe the business data from a logical perspective and not dependent on some actual system. This mode may be comprised of a canonical class model of the business entities and their corresponding relationships and the semantics, syntaxes and constraints of a superset of business attributes. These high level data models defining data structures often exclude class methods but in most cases the methods may be summarized into one business data object which is responsible for managing the structure.

Data structures also include refers to the way that data are stored in terms of relational tables and columns as well as how they can be converted into objects in object-oriented classes and how they can be structure with XML tags.

It is important to be very clear about the data structure because today’s information technology advancement, there can be a dozen of ways to implement a single of database tables such as an architecture where some of the rows may be on a computer in the United States while the other rows may be in New Zealand.

Data structures also pertain to the relationships between conceptual entities and the real data objects of the information system.

A metadata is basically any data about another data and they are very useful in facilitating a better understanding, management and use of data.  Therefore a common practice of many data warehouse implementations to include a metadata warehouse to enhance the performance of the whole system.

A metadata warehouses usually act as the interface in the data exchange between the data warehouse and the business intelligence. Since the metadata warehouse does not really contain the full data but just a description, the data structure is contained in the data component. The data components are very useful in data warehouses implemented in distributed heterogeneous environments.