An Active Metadata Warehouse is a repository of Metadata to help speed up data reporting and analyses from an active data warehouse. In its most simple definition, a Metadata is data describing data.
Many companies have spent years and billions of dollars trying to cleanse, profile, extract, transform, load and aggregate many different types of data into a Data Warehouse so they could use it in any way they want such as generate reports which are very accurate and share the business common views.
Many tools which are user-friendly query software applications and near real-time updating have been spent money on. To some degree, they have data which are accurate, accessible and timely. According to a recent report from a survey, less than 40 percent say they have come up wit accurate automated reports. This was because there never was a Meta data warehouse.
Most companies use an Active Data Warehouse to capture transaction data from many different sources. Since millions of data of transaction data may be processed during any given second in any data warehouse, storage for data is commonly separated in other computer from the operational system of the company. This is to ensure optimal resource management of the Data Warehouse Server. Also having a separate Active Metadata Warehouse significantly speeds up searching, analyzing and reporting data from the data warehouse.
An example of a Metadata would be to describe data called "rnt3263". Looking at that word alone does not seem to make sense. So, the metadata for "rnt3263" may be "the user identification of a client.
Metadata has various advantages. In its most wide usage, it is useful if we want to speed up our searches. Search queries with metadata expedite the process especially in performing very complex filter operations. Many web application locally cache Metadata by automatically downloading them and thus improve speed files access and searches. Locally, Metadata can be associated with files as in the case of scanned documents. When the files are digitally stored, the user may open the file using a viewer application which reads the document key values and store in a Metadata Warehouse or any similar repository.
Bridging a semantic gap is one of the notable uses of Metadata. For example, a search engine may understand that Edison is an American Scientist and so when one queries on "American Scientists" it may provide a hyperlink to pages on Edison even if the keyword did not mention "Edison". This approach is called semantic web and Artificial Intelligence.
Even in Multimedia files, Metadata is also very useful. For example, they are used to optimize lossy compression algorithms as in a video uses Metadata to tell the computer a foreground from the background to achieve better compression rate without losing much quality as effect of the lossy compression algorithm.
Metadata can be stored both internally meaning it is found within the file itself or externally as in a separate file that points to the file it describes. Storing Metadata externally is more efficient for searching like in Database queries
There are two general types of Metadata, Structural or Control Metadata and Guide Metadata. The Structural Metadata is generally used in database systems such as columns, tables and indexes. On the other hand, Guide Metadata is used to help users in looking for specific things like Natural Language searches.