(U01)
www.btinternet.com/~adrian.larner/database
ADRIAN LARNER
DATABASE PAGE
Return to The Home Page
DATABASE
Database Theory with radical proposals for a simplification of Relational Theory and a new interpretation of database records
Database Papers at this site
Reviews of Database Books for The Computer Journal
DATABASE ...
This fragment of the Database Page is under development.
See the following:
Copyright © 2001 Adrian Larner. The author, identified as Adrian Larner, asserts all moral rights. Subject to the asserted moral rights, the author grants pro tem unrestricted permission for non-commercial copying of all materials on this site for academic purposes.
The decorative image of a key (cc004239.gif) used on pages in this website was obtained from IMSI's MasterClips/MasterPhotos© Collection, 1895 Francisco Blvd East, San Rafael, CA 94901-5506, USA.
ADRIAN LARNER
DATABASE PAPERS
Return to The Home Page
Database Theory with radical proposals for a simplification of Relational Theory and a new interpretation of database records
PAPERS
Database Papers at this site
Reviews of Database Books for The Computer Journal
DATABASE PAPERS ...
Download
A New Model of Data
in rtf format (Word for Windows compatible)
A new model of data (excluding “real world” interpretation) is proposed. Informally it is close to the relational model, but – it is claimed – formally simpler, more powerful in expression, and conducive to greater data structure independence. The model amounts to a theory of records (effectively tuples) at a single level of aggregation (in contrast to the two or three levels of the relational theory). After the formal development of the theory a data manipulation language based on it is described, including a number of “outer” operations more tractable than relational outer joins. A treatment of “kept” (base, prime) data is described, which avoids some of the (admittedly slight) complexities of the relational approach. In any first order theory (lacking sets) the definition of ancestral predicates (transitive closures) can present problems: two specific examples are solved, one of them being that of bills of material. The theory permits records with missing and multiple values (non-uniform and non-flat records): justifications are advanced, and operators to apply restriction conditions to such records are defined.
Download
Nulls – A Socratic Enquiry
in rtf format (Word for Windows compatible)
An attempt is made to solve the major problems caused by nulls. The question, “What is a null?”, is posed, and the answer given is that: nulls are values (or, at least, there is no advantage to be gained by denying that nulls are values); it is not their meanings (e.g. unknown) that make them null; to define a column it is necessary to associate with it a criterion of identity that is the interpretation of “=” applied to its intended (“proper”) values; and a null is a value permitted in a column but not, under that criterion of identity, equal to anything. It is shown how nulls, under this definition, affect Natural Joins and primary keys (Entity Integrity). The handling of multiple nulls is discussed, as are (briefly and dismissively) default values, many valued logics, and nulls in Outer Joins.
Download
A New Foundation
for the ER Model
in rtf format (Word for Windows compatible)
A formal theory, based on first order logic, is proposed as an interpretive foundation for the Entity-Relationship (ER) model. Under the proposed interpretation, records are construed as existence assertions in which the kinds of entities asserted to exist are specified and/or their relationship is stated; in addition, a number of identities (one defining each attribute) are asserted of those entities. This theory allows an interpretation of derived records to be formally concluded from base records according to logical inference rules associated with data manipulation operations. In contrast with the Relational interpretation theory (the Domain Relational Calculus), the derivations are safe (false inferences – join traps – are avoided). The theory provides clarifications and explanations in a number of areas of ER theory, including entity and attribute definitions; meta-theoretical definitions, including those of “entity” and “attribute”; higher normalisation; entity dependence (referential integrity); and the teaching of ER analysis of data.
Download
A New Interpretation of Data
in rtf format (Word for Windows compatible)
The notion of data objects (records) modelling real world objects is examined and rejected in favour of interpretation of data objects as utterances. The weaknesses of the entity/relationship model are exposed, specifically its failure to interpret data manipulations. Next examined is the “classical” interpretation of records (including tuples in relations) as propositions in the language of the first order logic. This is shown to be too restrictive for the data that must be kept, largely because of its constraints on identity of values. A less constrained – relative – identity is proposed, and a new formal interpretation advanced and briefly examined.
Download
Object Oriented; Database:
Choose One Only
in rtf format (Word for Windows compatible)
Object-Oriented (OO) systems are inherently data structure dependent. Database (DB) systems are intended to be data structure independent. So to attempt to specify an OODB is vain. A simplified version of the Relational model provides data structure independence, and avoids the most telling criticism of that model: the lack of orthogonality of its structures.
Download
Objects and Entities
in rtf format (Word for Windows compatible)
A teaching paper:
If you are studying database specification
and object oriented analysis and design,
you will no doubt have observed that the entities and relationships shown
on entity/relationship diagrams seem to be very like
the objects and associations in class diagrams.
How are they related? What are their differences?
A teaching paper:
If you have a notation to use,
whether to read or to write, you must make sure that you understand it.
To understand it is to be able to translate it into English if reading it,
or to translate from English into the notation if you are writing it....
Let us begin with a symbol used in most data diagrams, [a] rectangle ... named PERSON.
What does it mean?
It represents a kind of record, a person record.
When we say that it represents a kind of record,
we mean that it represents each record of that kind.
In much the same way, the name “person” designates (that is, names)
each thing of a certain kind in the real world.
There is, of course, a connection between the real world object,
or entity (a person), and a person record in a file or database....