OpenCogPrime:FromSMEPHToOCPRevised

From OpenCog
(Redirected from SMEPH)
Jump to: navigation, search

From SMEPH to OCP

The SMEPH framework can be used to describe any intelligent system, including any AI system or a human brain. However, it can also be used as a guide for constructing particular AGI systems, such as OpenCogPrime.

OCP represents knowledge internally using a hypergraph data structure that involves nodes and links similar to SMEPH's edges and vertices. However, OCP's vocabulary of node and link types is richer than SMEPH's, and the semantics of its nodes and links are different than that of SMEPH's edges and vertices. For instance, OCP has node types called ConceptNode and SchemaNode, but also others like PredicateNode and various types of PerceptNodes. A OCP ConceptNode will not generally represent a SMEPH Concept edge, because it's rare that OCP's response to a situation will consist solely of activating a single ConceptNode. Rather, the Concept edges in the derived hypergraph of a OCP system will generally correspond to fuzzy sets of OCP nodes and links.

The term Map is used in OCP to refer to a fuzzy set of nodes and links that corresponds to a SMEPH concept or schema; and there is a typology of OCP maps, to be briefly discussed below. Often it happens that a particular OCP node will serve as the center of a map, so that e.g. the Concept edge denoting cat will consist of a number of nodes and links roughly centered around a ConceptNode that is linked to the WordNode cat. But this is not guaranteed — some OCP maps are more diffuse than this with no particular center.

Somewhat similarly, the key SMEPH dynamics are represented explicitly in OCP: probabilistic reasoning is carried out via explicit application of PLN on the OCP hypergraph, evolutionary learning is carried out via application of the MOSES optimization algorithm, and attention allocation is carried out via a combination of inference and evolutionary pattern mining. But the SMEPH dynamics also occur implicitly in OCP: emergent maps are reasoned on probabilistically as an indirect consequence of node-and-link level PLN activity; maps evolve as a consequence of the coordinated whole of OCP dynamics; and attention shifts between maps according to complex emergent dynamics.

A SMEPH-based intelligence such as a OCP system will also have a derived hypergraph, which will not be identical to the hypergraph it uses for explicit knowledge representation. However, an interesting feedback loop arises here, in that the intelligence's self-study will generally lead it to recognize large portions of its derived hypergraph as patterns in itself, and then embody these patterns within its concretely implemented knowledge hypergraph. This is closely related to the Cognitive Equation phenomenon described on the SMEPH page, in which an intelligent system continually recognizes patterns in itself and embodies these patterns in its own basic structure (so that new patterns may more easily emerge from them).