да, что-то не подумал... с английского на русский
фрагмент текста:
Abstract
The work described in this article arises from two needs. First, there is still a need for providing more sophisticated
database systems than just relational ones. Secondly, there is a growing need for distributed databases. These needs are
addressed by fragmenting schemata of a generic object data model and providing an architecture for its implementation.
Key features of the architecture are the use of abstract communicating agents to realise database transactions and queries,
the use of a remote object call mechanism to enable remote agents to communicate with one another, and the use of multilevel
transactions. Linguistic reflection is used to map database schemata to the level of the agents. Transparency for the
users is achieved by using dialogue objects, which are extended views on the database.
2005 Elsevier B.V. All rights reserved.
1. Introduction
The amount of data that is collected and stored in databases is steadily growing. In many cases the structure
of these data is not adequately reflected in the database schema and query languages. E.g. many databases
containing important biological data are organised more or less as flat files. In addition, the centralisation
of data in a big central database becomes more and more a handicap for efficient data processing. As data
arises at various locations and will be used at various other locations as well, the need for data distribution
increases.
As to the structure of the data, the research in the last decade has investigated complex values (i.e. data
constructed by various type constructors) and references/links between data – which in fact lead to infinite,
yet finitely representable structures.
