OK systems implementation

Knowledge Engineering and Robotics Lab

The computer implementation of such a system can be described in more precise terms. Organic Knowledge (OK) systems are ICT systems incorporating human expertise. One would be tempted to describe them as Expert Systems (ES) “on steroids” transforming them into Knowledge Systems (KS). If at all, they are an ecological system of many different and sometimes contradictory experts, called organs. But in reality the OK systems are so much more in almost​ every aspect that it is more correct to say they are the fulfillment of the ES vision. Other spiritual ancestors of OK system are Turing’s “child programs” and Minsky’s learning, evolving and non-algorithmic Society of Mind proposals.

 

​OK system is:

• evolving
• learning
• organized
• distributed
• dialectical
• very big knowledge base
  
  

Each organ is simulating an independent expert, and includes:

• knowledge base (data, meta-data and procedures)
• feedback apparatus:
  • knowledge acquisition mechanism (interfaces and communication)
  • learning mechanism (inference of new knowledge and processing)
  • evolution mechanism (creating and changing organs in view of the new knowledge)
• interfaces:
  • environment (local)
    • subjective (user)
    • objective
  • communication (network)
    • with other organs (o2o)
    • with remote servers
    • with remote users (p2p)
    • with remote resources
• execution (proactive)

 

 

The Gestalt-Multiplex-Layering (GML) combination described entails: gestalt – a deeper model of the expert knowledge and reasoning process; multiplicity – simultaneous use and cooperation of different and conflicting approaches; layering – use of a hierarchy of independent layers of control and processing, through which the input and intermediate results are propagated. The independence of each layer enables implementation of different approaches at different layers. The hierarchical layering of control and abstraction of lower by upper layers enables the cooperation and solution of contradictions arising from the use of a variety of different approaches. In very broad, plain terms, at each layer there is a small knowledge system controlling, generalizing and inducing the cooperation of different approaches in a larger knowledge system of the next layer.

 

Gesta​lt

Gestalt is the skeleton, the deeper model, the concept, the meta-model of the lower layer, abstracting, generalizing, controlling and interfacing it, and mediating between the lower layer, the upper layers and the user.

KS are frequently built around such approaches as ontology, schema, meta-model controlling and coordinating a multi-agent KS; tentative designs, templates; scenario generation; knowledge acquisition filtered by models; story model; fuzzy model; active behavioral database of goals and rules coordinating the knowledge database; declarative and executable object-oriented model; data structure at object level; frame templates; interaction manager; knowledge model used by agents to manage the others.
Gestalt has at least three facets: the declarative: the data and knowledge, the static aspect of knowledge; the procedural: the reasoning models, the data processing techniques, the inference engines, the dynamic aspect of knowledge; organizational: the interaction control of the multiplicity at the lower level, conflict resolution, user interface control, user involvement, feedback, the integration manager.
In general, the gestalt will involve a different approach than the lower level and include more than one component. Architecturally it is a quite complex structure. It can be viewed as a small knowledge system at the heart of and controlling the larger one. In a multi-agent society, it’s the ruler, the governing ideology and body.

 

Multiplicity

The basic commonplace of: “two heads are better than one” is validated by such hard science approaches as dialectics, to become a central component in the basic approach. The cognitive science teaches us that the integration of a variety of different techniques in a soft approach has immense advantages.
Many KS have been integrating more than one basic method: multi-agent society of different behavior patterns and different roles; integrating inductive decision trees and neural networks; multiple input channels and methods; CBR with RBR discrepancies solution; three independent models of knowledge representation and inference; integrating CBR, neural networks and discriminant analysis; blackboard method of integrating multiple experts (sources of knowledge); integrating KS with DB approaches; integration of various rule paradigms into a single KBS; a hybrid neuro-fuzzy reasoning; voting over multiple different learners; multi-knowledge systems combining such different approaches as extensional and intentional; integration of conflicting schemas; integrating different analytical decision model; integrating semantic expressions with examples; combination of relational and object-oriented paradigms.
As a guideline, the more different, even conflicting techniques are used, the wider and deeper view of the solution will be achieved. But this is too general and there arises the need for a more restrictive and precise method of integration
First option is merging into compromise. While compromise loses some of the positive features of each ingredient, a price paid for one monolithic consistent approach, merging recruits all, as contradictory as they may be, and the larger the diversity, the softer and fuzzier the result.
This requires such components built into the system as conflict resolution, control manager of interaction, data flow and resource allocation, user transparency and involvement, priority indexing of approaches. These are among the components of our gestalt, deep model driven, structural, hierarchical, layered architecture paradigm, which transforms the chaotic multiplicity into a well behaved, disciplined one.

 

Layered Approach

The layered approach proved itself very useful in such different areas as computer networks (​the reigning seven layers model), operating systems (e.g. the UNIX kernel-shells model), compilers (the three layered: lexical, syntactical and semantic model), client-server, front-end back-end approach. Among the many advantages are: independence, structuring, error protection, abstraction, precise interaction model, modularity, transparency, portability.
In the KS domain it has been widely used: structured libraries of behavior; client-server approach; hierarchical cases and domain specific indexing; two-level model with kernel and coordinating module; three layers: semantic, syntactic and lexical for structured processing; a hierarchical architecture modeling and inference; an object-oriented organizational layering; intentional layer over extentional; layering by generalization of schemas; hierarchical structuring of models; dual hierarchy, by structure and logic of data; three layers structured by human-computer interaction; a layered agents society.
In this approach the layering permits structured abstraction, conflict resolution by a higher level, different approaches at different layers, control and management, changeability and user involvement.

 

OK systems vs. other KS​

The essence of the approach is it’s softness, a very popular feature in many KS and general AI: fuzzy rules and terms; fuzzy logic in imprecise language systems; fuzzy qualitative constrains; fuzzy analysis; fuzzy inference integrated with neural network; fuzzy logic modeling.
It should be emphasized that though the layering concept has a linear structure connotation, the gestalt component certainly enables a much more complex architecture. The gestalt makes the structure dynamic rather than static, it actively and intelligently intermediates at each layer and between layers and directs the intermediate results. It’s rather a network of many possible interconnections and interactions between the different components of the system. The gestalt chooses at each stage the next path to be taken. The choice defines the layering, i.e. the sequence of nodes in certain order, of data processing, along the chosen path.
Such a distributed approach was adopted in multi-agent societies; multi-agent systems; distributed multi-agent environments.
The KS could be seen as a model of multi-agent society with one, very clever, best-connected agent, as the ruler.