Re: Latest version of glossary

"JOG" <jog_at_cs.nott.ac.uk> wrote:

>"The
>Entity-Relationship Model-Toward a Unified View of Data", 1976, P.
>Chen., his definition is:

. . .

>2 things here struck me as important. First, Chen writes that entities
>only exist 'in our minds', and as such are arbitrary constructions - I
>agree with him that this is an important part of any definition.

Or he has in mind what he means, quoted or otherwise:

"A specific person, company, or event is an example of an entity." So as person is one of his entity sets/types, so father, so son, so male, so female, would be one of his entities. If there were an entity set for, ancestors, if a parent were included then the same father which is an element of Person, who also be a member of, Ancestors.

But for entity, he admits, in the footnote, that it's not clearly defined.

His reason is explained in the introduction, as he says what problems he hoped to address with such a scheme, including:

"The relational model is based on relational theory and can achieve a high degree of data independence, but it may lose some important semantic information about the real world"

>Second, it appears that he ties the definition in very heavily in with
>an "entity set". From this he seems to be defining an entity as: "An
>element of an entity set". Good, clear and simple definition that ;)

No more so than, entity.

But you caught that:

>This would of course passes the buck to defining what an entity set is
>- but I believe that's an easier task, especially with Chen's reference
>to test predicates and common properties of elements of those sets.
>Opinions?

The predicate would be something like a dimensional constraint, whether a phone number if formatted as such, I would imagine. Things like that.

>Now I found another interesting contribution wtih Edward Yourdon, who
>describes E/R in his work Modern Structured Analysis, (Prentice Hall
>1989) and defines the concept of Entity as having three properties:

> 1. Each representation of an entity can uniquely be identified
> 2. Each representation of an entity is playing an important role in
>the system it lives in. (it has to have a reason to be there)
> 3. Each representation of an entity can be described by one or more
>attributes (data-elements, like name, age, quantity)

The term, entity, is used in various contexts, and seems vague. So does the word, fact. But a datum is not ambiguous. A scalar remains a scalar. And a string of characters remains a string of characters.

But in order to operate on a set of 'things', one must refer to these, 'things', as . . . things, basically, entities, what have you. And so on. Received on Sat Feb 25 2006 - 06:32:52 CET