Re: Weak Entity

From: Jan.Hidders <hidders_at_hcoss.uia.ac.be>
Date: 14 Oct 2002 11:29:46 +0200
Message-ID: <3daa8e8a$1_at_news.uia.ac.be>


In article <aod8v1$5qi$1_at_news.ctimail.com>, Peter Kwok <peterkwok628_at_yahoo.com> wrote:
>> >In ER diagram, should weak entity be related to strong entity in 1-N
>> >relationship?
>> >i.e.
>> >Correct:
>> >Strong Entity----------(1)-Relate-(N)----------Weak Entity
>> >
>> >Wrong:
>> >Strong Entity----------(M)-Relate-(N)----------Weak Entity
>>
>> No.
>>
>> For example:
>>
>> Element --(M)-Is_contained_in-(N)-- Non-empty_Set
>>
>> Clearly non-empty sets cannot exist without their elements and they have
>> no attributes to form a key.
>>
>> However, some definitions of the ER model do have this restriction, so it
>> also depends a bit on what book you are using and how smart your
>> instructor is. :-)
>
>However, how can weak NON_EMPTY_SET be uniquely identified by relating to
>its partner-ELEMENT under such many-many relationship?

In the ER model or in the relational model you map it to? The usual mapping is that you include the key of the strong entity that the weak entity depends upon with the relation for the weak entity. And that indeed does not work if you have a M-N relationship, so you do have a problem in your mapping then. But this is a problem of the mapping, not of the relational model or the ER model.

  • Jan Hidders

>//Peter
>
>
Received on Mon Oct 14 2002 - 11:29:46 CEST

Original text of this message