Traditionally (and inevitably) knowledge gets represented as a top down branching tree structure where there is a 1 to many downwards relationship and a many to 1 upward relationship and an assumption of mutual exclusivity between the branches. (i.e. - a category that apppaers in 1 branch cannot appear in another branch , and has no relation to the other branch beyond the point where they have a common fork at the higher level.
In human terms - a cousin cannot be your brother - (or at least, shouldn't be.)
There are many well-know trees out there that attempt to classify all knowledge into tree structures - including:
The Dewey Decimal System
The Library of Congress (LCSH) number (compared)
ISBN
dmoz
NAICS
the Tree of Life (i.e. the Phylogenetic tree)
and many, many more
the more successful ones tend to limit their range
----------------------------
in Computer Science these are expressed as radix trees