P. Atzeni and P. P. Chen. Completeness of Query Languages for the Entity-Relationship Model. In P. P. Chen, editor, Proceedings of the Second International Conference on EntityRelationship Approach, pages 111--123, October 1981.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....associated commentaries are available via anonymous ftp from FTP.cs.arizona. edu:tsql tsql2 tsql2.final. 1 the calculus is said to be complete. Original work along these lines was done by Codd for relational databases [Cod72] There have been generalizations for entity relationship databases [AC81] and for temporal databases [TC90] One inherent problem with these approaches is the degree of appropriateness of the calculus that is used as a metric. There is no guarantee that the calculus captures all reasonable queries. For example, it has been shown [AU79] that first order relational ....

P. Atzeni and P. P. Chen. Completeness of Query Languages for the Entity-Relationship Model. In P. P. Chen, editor, Proceedings of the Second International Conference on EntityRelationship Approach, pages 111--123, October 1981.

....(usually first order relational calculus) as a metric. Any language having at least the expressive power of the calculus is said to be complete. Original work along these lines was done by Codd for relational databases [11] There have been generalizations for entity relationship databases [1] and for temporal databases [30] One inherent problem with these approaches is the degree of appropriateness of the calculus that is used as a metric. There is no guarantee that the calculus captures all reasonable queries. For example, it has been shown [3] that first order relational calculus ....

P. Atzeni and P. P. Chen. Completeness of Query Languages for the EntityRelationship Model. In P. P. Chen, editor, Proceedings of the Second International Conference on Entity-Relationship Approach, pages 111--123, October 1981.

....from three basic assumptions : operands and result are entity types. Each operator is designed to manipulate one (or more) entity type(s) and to build the result as a new entity type, added to the existing database. Thus, the result may serve as an operand for a subsequent operator. Unlike [Atzeni 83, Chen 84] we do not provide query operators on relationship types nor on attributes. The rationale for this is both to be consistent with the primary role of entity types in an ERC database and to keep the algebra as simple as possible. This is also conform to the object oriented paradigm, ....

....as a single object type through an algebraic expression mainly based on r joins. 4.2 Product A product operator is needed to dynamically connect two entity types for which the desired link does not exist in the actual schema. In this way, the set of possible queries is not restricted (as in [Atzeni 83] for instance) to those foreseen at the design stage. Product joins every occurrence of the first entity type with (the multiset of) all occurrences of the second entity type. As in the relational algebra, the result of a product, if isolated, is not very meaningful. It becomes meaningful when ....

P. ATZENI, P. P. CHEN: "Completeness of query languages for the entity-relationship model", in Entity-relationship approach to information modeling and analysis, Chen P.P. Ed., North Holland, 1983

....as a practical tool, the theoretical framework for the ER approach has received relatively little attention from the research community. Fundamental basis for ER data manipulation are poorly defined. Only a few attempts to define an ER algebra [Markowitz 83] Campbell 85] or an ER calculus [Atzeni 83] Chen 84] Hohenstein 88] have been reported, and, in our opinion, they fail to be satisfactory. Both based on a rather primitive version of the ER model, the two algebras merely provide a reshaping of the relational algebra. They are not appropriate for taking into account the significant ....

....version of the ER model, the two algebras merely provide a reshaping of the relational algebra. They are not appropriate for taking into account the significant features of modern, more sophisticated ER models (where, for instance, complex attribute structures are allowed) The calculus in [Atzeni 83] is too poor to be of any use, while the one proposed in [Chen 84] only supports a primitive ER model. The only significant work in this area has been recently reported: Hohenstein 88] presents a calculus, but no algebra, for an extended ER model, named EER. The model, for which the authors ....

[Article contains additional citation context not shown here]

P. Atzeni, P.P. Chen: "Completeness of Query Languages for the Entity-Relationship Data Model", in Entity-Relationship Approach to Information Modeling and Analysis, P.P. Chen Ed., North Holland, 1983

....of such queries is: Find the global amount of salaries of the employees ; answering this query requires the ability to recognize numbers in the database, to add and to produce a number as output. 4. E R calculus queries: they correspond to the class of queries obtained by using the E R calculus [4], a calculus similar to the relational calculus defined in [11] the only difference is that atoms containing references to two or more different object sets (either entity sets or relationship sets) are allowed only if the object sets are explicitly related to the schema of interest. This ....

ATZENI, P. AND CHEN, P.P. 1981. Completeness of Query Languages for the Entity Relationship Model, Proc. of the 2nd Intl. Conference on the Entity Relationship Approach to Information Modelling and Analysis (P.P. Chen ed.), Washington D.C., North Holland.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC