Different classes of recursive queries in the relational databases are identified. It is shown that existing proposals to extend the relational query languages are either not powerful enough to express queries in many of these classes or use nonfirst normal form constructs. RQL, a recursive database query language that can be used to express recursive queries on all the classes identified, is presented. RQL is based on the relational algebra. In addition to functions that correspond to the standard and extended relational algebra operators, RQL supports functions required to express general recursive queries. The elements of RQL and the ways in which they are used to formulate complicated, but useful, recursive queries are described. The effects of the extensions embodied in RQL on the termination of recursive query evaluation are discussed.>

In knowledge-based system (KBS) applications where the number of facts in the knowledge base is very large, the amount of information that is needed for effective explanation support can become too voluminous to be stored in main memory. The authors present an approach to modeling and managing the information needed for explanation rising a relational database. It is shown how different types of explanation can then be produced by appropriate queries on this database. The authors formulate representative queries for some major types of explanation using ESQL, an extension of structured query language (SQL).>

It is shown that by simply extending the relational data model to support image domains, the model becomes rich enough for many advanced applications. Extensions to the relational algebra and the relational calculus are developed to exploit the semantics of image domains. The extended calculus is incorporated in the query language ESQL, which is a strict superset of the structured query language (SQL). A technique to implement ESQL using a preprocessor to SQL is shown. A noteworthy observation is that it is possible to incorporate multirelations without any modifications to the storage structures and data definition language of SQL. It is also possible to support queries of multirelations in ESQL without modifying the features of SQL in any way as far as traditional data manipulation is concerned.>

We present an analytical formula for estimating the cardinality of the projection on certain attributes of a subset of a relation in a relational database. This formula takes into account a priori knowledge of the semantics of the real-world objects and relationships that the database is intended to represent. Experimental testing of the formula shows that it has an acceptably low percentage error, and that its worst-case error is smaller than the best-known formula. Furthermore, the formula presented here has the advantage that it does not require a scan of the relation.