Generally speaking, a database is a collection of data fulfilling certain conditions: for example, the data has to be managed to ensure its integrity and quality, it should allow shared access by a community of users, and should also have a schema or support a query language.
More specifically, termbase (the short form for terminology database - TDB) has been defined as "the collection of information on a term or concept in a structured, electronically readable way combined with a terminology management system". Though it is often used as synonym of termbank, there are some differences. According to terminologists, for example, termbanks include the organizational environment and terminology databases do not (cf. Maslanko, 2004:9).
According to their structures, databases can be relational, hierarchical or object-oriented.
- In relational databases information is stored in tables which are related to each other by "primary keys". These keys allow locating and processing the related information that the user wants to. The standard language for relational databases is the Structured Query Language (SQL). Some examples of relational database systems are Microsoft Access, Oracle, Lotus Approach, etc.
- Hierarchical databases (or networks) enable the user to create hierarchies of data inputs but the searching functionality for specified data is limited. For example, to obtain data it is necessary to navigate from the "starting" or "entry points".
- A promising approach are the so called object-oriented databases (OODB). Data records stored in relational tables are represented as objects, which combined with database capabilities produces a database ideal for high performance and complex data.