Reputation: 21
Why is a 3 binary relationship relationship representation more general than one ternary relationship representation?
Upvotes: 2
Views: 7573
Reputation: 751
Because with 3 binary relationships each involved entity is related separately with each one of the others two.
Assumed that ternary relationship in essential only when you have a many-to-many-to-many relationship (otherwise you can rewrite it adding a relational entity linked with binary relationships to the others three), let's take an example:
Suppose that you have the entities A, B and C, and three relationships that link the entities "like a triangle". Now, suppose that set of tuples a1 is related to the sets b1 and c1. Using 3 binary relationships, it's not requested that b1 and c1 are related. for example, b1 can be related to c2, that can partially overlap c1 or be completely disjoined.
With a ternary relationship, instead, b1 must be related to c1.
So you can see that with 3 binary relationships you can have much more combinations.
Upvotes: 4
Reputation: 27424
In general one defines properties of relationships, like cardinality (only on object of a set can partecipate to the relationships or more than one, for istance), and totality/partiality (all the objects of the set must partecipate / can partecipate to the relationship).
So, limiting to these two properties with the two possibilities, in a ternary relation one has 12 possible combinations of values for the properties (3 x 2 for cardinality and 3 x 2 for totality).
If a ternary relationship is represented instead with 3 binary relationships, there are, for each endpoint of a relationship (there are six of them) 4 possibilities (single or multiple, tutal or partial), so in total one can have 24 possible combinations, that this a number of possible specifications which is the double with respect to that of the ternary relationship.
Upvotes: 0