Research into Declarative, Object-Oriented Languages for Programming Data and Knowledge

点击翻译按钮获取中文摘要

Relational query languages, such as SQL, were the early success stories of declarative programming. However, as new applications areas have developed, relational systems have been found wanting. Their limitations--both structural and language related--have led databases and deductive databases. Object-oriented databases (abbr., O-O) are effective when the structure of the data is well- known and the data access paths and retrieval methods can be determined at system development time. Deductive DBMS, based on logic, are appropriate in less well-defined situations; situations where one does not know before-hand all the queries that will be asked. Flexibility is extremely important for modern DBMS, especially when the user lacks detailed knowledge about the structure of the application domain. For this reason, logic-base languages have traditionally been used whenever intelligent processing of data was required. On the other hand, flat data structures inherent in such languages are ill-suited for many scientific applications, CAD/CAM design, modeling agile manufacturing processing, etc., where data has complex and known internal structure.