Modern Relational Database Management Systems can be configured to collect data at certain events. In a business, some of the events are sale, order, deposit, pay, request and many more. Events are really business activities in a company. Big companies, in order to gain competitive advantage, invest heavily on setting up business data warehouses and employ software based business intelligence systems.
In today’s modern data warehouses, a popular data modeling used is the event-entity-relationship model (EVER). This Data Model is general purpose modeling scheme used to support both the business general schema structure and multidimensional scheme of a particular business set-up for serving specific information needs.
In a more technical description, an event can be described as any unique or immutable representation of an occurrence of an atomic process phenomenon. Another term, an even set, refers to a named set of events. An entity, on the other hand, refers to an immutable representation of the uniqueness of a mutable phenomenon.
For example, in library warehouse, librarians, books and borrowers can be the entities while borrowing, returning and putting penalties can be the events. In the EVER, the relationships between entity and events are taken cared of by a relational database management system so that any specific information needed may be acquired with a simple query.
Events can be triggered manually or automatically. For example, in a business company, whenever the sales department generates a single sales and the transaction is recorded in the database, this is already an event that is manually triggered. But this is not always the case.
In fact, a multitude of other events are automatically triggered. For example, if a company has several branches and need to do batch processing of transferring sales records from different branches across several countries, each of the country databases may be configured to transfer records synchronized at a specific time.
The database administrators of each database may configure the system to automatically load the sales records to the central database so the systems around the world can automatically trigger the event of transferring data from one country server to a central database.
Recording of collection is extremely important. This is just one aspect of the logs of transactions. When the logs contain collection times of events, it may be a lot easier to troubleshoot should problems arise during the event.
For instance, if the central server breaks at a particular point and the database administration only noticed the problem a few minutes later, he can trace the Collection Time and check which among the branch databases caused the problem or which among the data cause the problem. If there was no Collection Time record, it may become hard to trace the history of transactions.
Many data warehouses also employ a data modeling called Entity Life Histories – Notation. In this technique, an entity’s events are recorded. For an entity may be a customer name. A customer’s events include birth events, death events, middle life events and death events.
Each event can affect the entity in some ways during its life in the whole system. The same events can cause the entity to be created, removed or modified in the database. And in each modification, removal or re-addition, collection time of data is always recorded by the database.
Business intelligence use Collection Time as an integral part of its statistical analysis process. With Collection Time, specific trends can be spotted and appropriate innovations can be created by the business.
For example, the business intelligence in an airline senses that the month of June is a period where many travelers prefer flying with holiday tour packages, then they may need to form better alliance with the hotel industry and come up with enticing promos.