OntologyStream Inc

Technical note on n-ary relationships and event valence

Monday, December 10, 2001

In Figure 1 we show a relationship that is ternary (left) and an atom what has three affordances.

Figure 1

The notion of affordance is similar to a notion of chemical valance, in that affordances allow different types of aggregation of atoms into compounds.

We start with what we call an analytic conjecture.

Figure 2

This establishes a binary relationship between values in one data base column (these values are abstract atoms) and a value in a different column. The first column values link second column values.

The analytic conjecture could be a determination of a 4nary relationship of the type shown in Figure 3. This will give a different “resolution” on essentially the same information. Like a fish net, the larger the mesh the more easily the net is to pull, but the smaller fish get away.

Figure 3

The SLIP (Service Link analysis, Iterated scatter-gather, and Parcelation) takes two columns of data and converts this into two intimately related abstractions, “links” and “atoms”. These are abstractions in exactly the same way as the number “one” is an abstraction. Given any context the abstraction can be demonstrated immediately. There is no false information or incomplete information.

“Give me all links that are related to this particular atom and that is related to an second atom from a specific “bag” of atoms.” \

This is a query that produces a perfect recall with perfect precision. The science of data invariance is born (perhaps?)

OSI’s Referential Information Base (RIB) concept takes these two columns and orders the column values by the number base that accounts for all of the characters used in all of the values. Base 64 works for all text words. The column values are attached to a header that references back to the record number in the original data source. This is used for query purposes.

Figure 4

The two columns are now simply two geometric objects (lines or tensors of rank 0) where very fast search algorithms can be used to do set and category theory. The emergent computing process (a scatter gather process adopted from text data and n-gram analysis) is made fast enough to allow real time interaction without a super computer.

OSI’s roll out of SLIP Incident Management and Intrusion Detection Systems (IMIDS) is ready to be evaluated by industry management.