Data warehouse – definitions

Notions used by BAKK in the scope of data warehouse: data warehouse, data marts, data cubes, MIS (Management Information System), DSS (Decision Supporting System), BI (Business Intelligence), multidimensional data structures, dimensions, measures, hierarchies etc.

data warehouse

– separate database, integrated and centralized repository of cross-sectional data pertaining to numerous aspects of organization or company activities; organizes data according to business categories e.g. time, product, client; data often comes from various sources, including external ones, most often from operational or transactional databases.

Business Intelligence (BI)

– systems which enable detecting relations between data, offering more than passive data presentation for the sake of analysis; they are based on data repositories such as data warehouses.

Decision Support System (DSS)

– systems which enable dynamic queries to data repository, without the knowledge of SQL query language necessary.

Executive Information System (EIS)

– systems which enable dynamic queries to data repository, without the knowledge of SQL query language necessary, only within the scope of rigidly defined frames and conditions, which have been defined in query interface.

On Line Analytical Processing (OLAP)

– technology of processing, storing and providing access to data in the manner which is optimized for analytical application.

On Line Transaction Processing (OLTP)

– technology of processing, storing and providing access to data in the manner which is optimized for transactional application – most often for recording economic events within a company.

multidimensional data structures

– data structures which – in logical meaning – enable saving/reflecting multidimensional space. In other words, to assign localization attributes for events to each of n dimensions. To be differentiated from multidimensional databases.

relational data structures

– data structures which – in logical meaning – enable saving/reflecting relations of links between objects. Subset of relational data structures is constituted by multidimensional structures, which have been constructed on the basis of properly built relations.

multidimensional databases

– databases operating on notions of ‘measure’, ‘dimension’ and ‘hierarchy’ in isolation from their technical manner of recording.

relational databases

– databases operating on notions of tables and relations between tables.

measure

– data warehouse object, the role of which is to register and store presented values (most often numeric) e.g. number of sold products, storehouse stock list, but also text e.g. commentary (it is a subject for discussion whether it is a measure or not, because non-numeric measures cannot undergo many operations traditionally performed on measures e.g. aggregating).

dimension

– data warehouse object, the role of which is to organize data recorded within measures so as to allow for its analysis – according to one aspect e.g. dimension of time, geography, product.