Ontology Modeling Distinguishing Parts From Kinds

Available is a simple Ontology:

* Mortal

* Close relative
* Commencement
* Tot

* Boy
* Young woman

* Femininity

* Staff

* Womanly

We read the Ontology like this:

Close relative "is a" Mortal

Commencement "is a" Mortal

Tot "is a" Mortal

Boy "is a" Tot "is a" Mortal

Young woman "is a" Tot "is a" Mortal

Staff "is a" Femininity

Womanly "is a" Femininity

If the "X is a Y" relation are logical (does the relationship make think logically in the real world?), after that the Ontology class hierarchy is impregnable.

Here's an example of an hearsay Ontology class hierarchy:

* Mortal

* Commencement

* Staff

This model asserts that "Staff is a Commencement". Or, if you were to enter this in a somewhat manager close air, "all males are fathers". Utterly for some, not true for others. So this is an ridiculous relation

Communicate are an assortment of ways to enter hierarchies.

A joint effortless hierarchy is:

* State-owned

* Phylum

* Sharing

* Warn

* Kin

* Lineage

* Species

This is a great use of a tree-like method to show that

Kingdoms defend Phylums

Phylums defend Sequence

Sequence defend Tips

etc.

In the same way, it shows that

Species are a part of Lineage,

and a Lineage is a part of Kin,

and a Kin is a part of an Warn

etc.

This tree method (however useful under an assortment of dispatch) is not a impregnable Ontology.

In an Ontology, we would blow your own horn to read this as:Species "is a" Lineage

Lineage "is a" Kin

Kin "is a" Warn

Warn "is a" Sharing

Sharing "is a" Phylum

Phylum "is a" KingdomAnd none of these statements are true. In the first Ontology model, a Commencement "is a" Mortal and a Close relative "is a" Mortal. But a Lineage "is not a" Species, and a Phylum "is not a" State-owned.

It is major to gossip relating parts and kinds:

A Phylum "is part of a" State-owned.

A Phylum "is a group of" State-owned.

A Close relative "is a part of a" Person.

A Close relative "is a group of a" Person.This is habitually the difference relating a Organization and an Ontology. A Organization doesn't need to gossip relating parts and kinds. An Ontology constraint make this delicacy.

Notice that as we edit "is a group of" to "is a", we can further edit "is part of" to "has a". So again, we can say "Warn "has a" Kin" but not "Kin "is a" Warn" References:

* http://www.w3.org/2001/sw/BestPractices/OEP/SimplePartWhole/

* http://www.cs.sjsu.edu/~pearce/modules/patterns/hierarchies/composite/partonomy.html