"In situation analysis (SA), an agent observing a scene receives information...Considering the logical connection between belief and knowledge, the challenge for the designer is to transform the raw, imprecise, conflictual and often paradoxical information received from the different sources into statements understandable by both man and machines."
Decision support through constraint propagation in collaborative distributed command and control
"In this paper we develop a conceptual model of the interdependences among plans that can be expected to emerge in a collaborative, distributed command and control center. The foundations of the model are the problem space representation of problem solving and analyses of the nature of constraints and their propagation and of the task of planning. The model has informed the development of a series of empirical studies of the propagation of constraints in a simulated command and control center. The CBFire microworld is the test bed for the studies. Analysis of the behavioral data captured by C3Fire would serve to inform the design of an intelligent interface for decision support in command and control that highlights constraints on action and facilitates human decision making"
"Research on dynamic systems was motivated partly because traditional IQ tests turned out to be weak predictors in non-academic environments (see Rigas & Brehmer, 1999, p. 45). According to their proponents, computer-simulated “microworlds ” seem to possess what is called “ecological validity”. Simulations of (simplified) industrial production (e.g., Moray, Lootsteen, & Pajak, 1986), medical systems (e.g., Gardner & Berry, 1995), or political processes (e.g., Dörner, 1987) have the appeal of bringing “real-world tasks” to the laboratory. Brehmer and Dörner (1993) argue that these scenarios escape both the narrow straits of the laboratory and the deep blue sea of the field study, because scenarios allow for a high degree of fidelity with respect to reality and at the same time allow systematic control of influential factors.
Subjects acting in these scenarios did indeed face a lot more tasks than in the IQ tests: (a) the complexity of the situation and (b) the connectivity between a huge number of variables forced the actors to reduce a large amount of information and anticipate side effects; (c) the dynamic nature of the problem situation required the prediction of future developments (a kind of planning) as well as long-term control of decision effects; (d) the intransparency (opaqueness) of the scenarios required the systematic collection of information; (e) the existence of multiple goals (polytely) required the careful elaboration of priorities and a balance between contradicting, conflicting goals."
Complex problem solving: the European perspective
Komplexes Problemlösen vor dem Hintergrund der Theorie finiter Automaten:
"The theory of finite state automata is presented as a useful tool for problem solving research. For investigations of how people interact with discrete dynamic systems the approach suggests hypotheses about system knowledge acquisition and representation, it serves to deduce knowledge measures, and it enables researchers to describe and to construct entire classes of discrete dynamic task environments. A number of experimental studies are described in order to illustrate the approach."
Goal specificity effects on hypothesis testing in problem solving:
"Previous research has found that having a nonspecific goal (NSG) leads to better problem solving and transfer than having a specific goal (SG). To distinguish between the various explanations of this effect requires direct evidence showing how a NSG affects a participant's behaviour. Therefore we collected verbal protocols from participants learning to control a linear system consisting of 3 outputs by manipulating 3 inputs. This system was simpler than the one we had used previously, so in Exp. 1 we generalized our earlier goal specificity findings to this system. In Exp. 2 protocol analysis confirmed our prediction (based on dual-space theories of problem solving) that NSG participants focused on hypothesis testing whereas SG participants focused on the goal. However, this difference only emerged over time. We also replicated the goal specificity effect on performance and showed that giving participants a hypothesis to test improved performance."
Quinn: Why so many big words. Here's the elevator pitch version: "In Situational Analysis -- people see s**t and try to figure it out."
Hamid: Not just people, machines too...Canada's defense force of the future?
Barwise and Perry
Situations, unlike worlds, are not complete in the sense that every proposition or its negation holds in a world. According to Situations and Attitudes, meaning is a relation between a discourse situation, a connective situation and a described situation. The original theory of Situations and Attitudes soon ran into foundational difficulties. A reformulation based on Peter Aczel's non-well-founded set theory was proposed by Barwise before this approach to the subject petered out in the early 1990s.
Barwise and Perry's system was a top-down approach which foundered on practical issues which were early identified by Angelika Kratzer and others. She subsequently developed a considerable body of theory bottom-up by addressing a variety of issues in the areas of context dependency in discourse and the syntax-semantics interface. Because of its practical nature and ongoing development this body of work "with possible situations as parts of possible worlds, now has much more influence than Barwise and Perry’s ideas".
Situation Theory and Situation Semantics:
“The world consists not just of objects, or of objects, properties and relations, but of objects having properties and standing in relations to one another. And there are parts of the world, clearly recognized (although not precisely individuated) in common sense and human language. These parts of the world are called situations. Events and episodes are situations in time, scenes are visually perceived situations, changes are sequences of situations, and facts are situations enriched (or polluted) by language.
Types and the type abstraction procedures provide a mechanism for capturing the fundamental process whereby a cognitive agent classifies the world. Constraints provide the situation theoretic mechanism that captures the way that agents make inferences and act in a rational fashion. Constraints are linkages between situation types. They may be natural laws, conventions, logical (i.e., analytic) rules, linguistic rules, empirical, law-like correspondences, etc. For example, humans and other agents are familiar with the constraint"
Lemuel: How refined is the measurement?
Hamid: Alternatively, how does the measurement refine?
Lemuel: I am referring to the measurement process as it has been shown in some other studies that the mathematical model is adjusted along quantum lines.
Lemuel: Hamid, how would one handle false signals? Say, a false missile launch? I am thinking along widely varying issues, sort of tricks, once you know the system.
Hamid: One approach to dealing with false signals or "Error-Correction" is a form of heirarchical redundancy which globally "sums-over" local-deviations from perfectly complementary attributive (state-syntax or topological-descriptive) mappings between objects having properties sharing mutual relations.
An ontology is a specification of a conceptualization.
The word "ontology" seems to generate a lot of controversy in discussions about AI. It has a long history in philosophy, in which it refers to the subject of existence. It is also often confused with epistemology, which is about knowledge and knowing.
In the context of knowledge sharing, I use the term ontology to mean a specification of a conceptualization. That is, an ontology is a description (like a formal specification of a program) of the concepts and relationships that can exist for an agent or a community of agents. This definition is consistent with the usage of ontology as set-of-concept-definitions, but more general. And it is certainly a different sense of the word than its use in philosophy.
What is important is what an ontology is for. My colleagues and I have been designing ontologies for the purpose of enabling knowledge sharing and reuse. In that context, an ontology is a specification used for making ontological commitments. The formal definition of ontological commitment is given below. For pragmetic reasons, we choose to write an ontology as a set of definitions of formal vocabulary. Although this isn't the only way to specify a conceptualization, it has some nice properties for knowledge sharing among AI software (e.g., semantics independent of reader and context). Practically, an ontological commitment is an agreement to use a vocabulary (i.e., ask queries and make assertions) in a way that is consistent (but not complete) with respect to the theory specified by an ontology. We build agents that commit to ontologies. We design ontologies so we can share knowledge with and among these agents.
Ontologies as a specification mechanism
A body of formally represented knowledge is based on a conceptualization: the objects, concepts, and other entities that are assumed to exist in some area of interest and the relationships that hold among them (Genesereth & Nilsson, 1987). A conceptualization is an abstract, simplified view of the world that we wish to represent for some purpose. Every knowledge base, knowledge-based system, or knowledge-level agent is committed to some conceptualization, explicitly or implicitly.
An ontology is an explicit specification of a conceptualization. The term is borrowed from philosophy, where an Ontology is a systematic account of Existence. For AI systems, what "exists" is that which can be represented. When the knowledge of a domain is represented in a declarative formalism, the set of objects that can be represented is called the universe of discourse. This set of objects, and the describable relationships among them, are reflected in the representational vocabulary with which a knowledge-based program represents knowledge. Thus, in the context of AI, we can describe the ontology of a program by defining a set of representational terms. In such an ontology, definitions associate the names of entities in the universe of discourse (e.g., classes, relations, functions, or other objects) with human-readable text describing what the names mean, and formal axioms that constrain the interpretation and well-formed use of these terms. Formally, an ontology is the statement of a logical theory.
We use common ontologies to describe ontological commitments for a set of agents so that they can communicate about a domain of discourse without necessarily operating on a globally shared theory. We say that an agent commits to an ontology if its observable actions are consistent with the definitions in the ontology. The idea of ontological commitments is based on the Knowledge-Level perspective (Newell, 1982) . The Knowledge Level is a level of description of the knowledge of an agent that is independent of the symbol-level representation used internally by the agent. Knowledge is attributed to agents by observing their actions; an agent "knows" something if it acts as if it had the information and is acting rationally to achieve its goals. The "actions" of agents---including knowledge base servers and knowledge-based systems--- can be seen through a tell and ask functional interface (Levesque, 1984) , where a client interacts with an agent by making logical assertions (tell), and posing queries (ask).
Pragmatically, a common ontology defines the vocabulary with which queries and assertions are exchanged among agents. Ontological commitments are agreements to use the shared vocabulary in a coherent and consistent manner. The agents sharing a vocabulary need not share a knowledge base; each knows things the other does not, and an agent that commits to an ontology is not required to answer all queries that can be formulated in the shared vocabulary.
In short, a commitment to a common ontology is a guarantee of consistency, but not completeness, with respect to queries and assertions using the vocabulary defined in the ontology.
 Ontologies are often equated with taxonomic hierarchies of classes, but class definitions, and the subsumption relation, but ontologies need not be limited to these forms. Ontologies are also not limited to conservative definitions, that is, definitions in the traditional logic sense that only introduce terminology and do not add any knowledge about the world (Enderton, 1972) . To specify a conceptualization one needs to state axioms that do constrain the possible interpretations for the defined terms."
"The first paper uses the channel logic of Barwise and Seligman to define ontology morphisms, which can be used to integrate ontologies. The second paper uses the local logics of channel theory and Chu spaces to formalize a duality between ontologies and their instantiantiations; this paper also cites my early work on using colimits (from category theory) for knowledge integration, and also includes some interesting examples of integrating knowledge in ontologies over different logics. The third paper discusses composition operations on ontologies languages, again using the local logics, and also discusses the akt editor for applying such operations. The fourth paper discusses a tool for evolving large distributed knowledge based on ontologies, making use of histories of transformations.
The first three papers are closely related to my work on "institutions," an abstract axiomization of the notion of "logical system," much of it with Rod Burstall of Edinburgh (see section 1.4). Local logics and Chu spaces are both special cases of institutions, and the duality is also a special case of the syntax-semantics duality that is built into institutions. Moreover, the ontology morphisms of the first paper are a special case of theory morphisms over an institution. (For the cognescenti, V-insitutions generalize Chu spaces, and were proposed for similar applications long before Chu spaces.) Local logics do not appear to allow a sufficiently strong distinction between the object level of ongologies and the meta level of ontology languages; this distinction is much clearer with institution theory, and also, it is know how to obtain much more powerful composition operations in that framework, because composition of parameterized software modules is one of its major applications."
Von Schweber Living Systems
"The name Synsyta stems from a term in cell biology, syncytium : Tissue in which multiple cells come together and operate as a single cell, as is the case with neurons and muscle cells. The central image to the right is a syncytium.
“It is ironic that gap junctions connect together neurons and glia, at least transiently, into a sort of reticular syncytium— Golgi’s idea overthrown by Cajal’s demonstration of discrete neurons and chemical synapses. Gap junction assemblies of transiently woven-together neurons have been termed “hyper-neurons” and suggested as a neural correlate of consciousness.” (Consciousness, the Brain, and Spacetime Geometry, Hameroff)
“As few as three gap junction connections per cortical neuron (with perhaps thousands of chemical synapses) to neighboring neurons and glia which in turn have gap junction connections elsewhere may permit spread of cytoplasmic quantum states throughout significant regions of the brain, weaving a widespread syncytium whose unified interior hosts a unified quantum state or field (Hameroff & Penrose, 1996; Woolf & Hameroff, 2001).”
A Syncytium, shown below, is a group of cells that come together and act as one cell. Neurons and muscle cells form syncytiums."
"In recent years gamma synchrony has indeed been shown to derive not from axonal spikes and axonal-dendritic synapses, but from post-synaptic activities of dendrites. Specifically, gamma synchrony/40 Hz is driven by networks of cortical in...ter-neurons connected by dendro-dendritic “electrotonic” gap junctions, windows between adjacent cells. Groups of neurons connected by gap junctions share a common membrane and fire synchronously, behaving (as Eric Kandel says) “like one giant neuron.” Gap junctions have long been recognized as prevalent and important in embryological brain development, but gradually diminish in number (and presumably importance) as the brain matures. Five years ago gap junctions were seen as irrelevant to cognition and consciousness. However more recently, relatively sparse gap junction networks in adult brain have been appreciated and shown to mediate gamma synchrony/40 Hz.1-11 Such networks are transient, coming and going like the wind (and Hebbian assemblies), as gap junctions form, open, close and reform elsewhere (regulated by intraneuronal activities). Therefore neurons (and glia) fused together by gap junctions form continually varying syncytia, or Hebbian “hyper-neurons” whose common membranes depolarize coherently and may span wide regions of cortex. (The internal cytoplasm of hyper-neurons is also continuous, prompting suggestions they may host brain-wide quantum states.) By virtue of their relation to gamma synchrony, gap junction hyper-neurons may be the long-sought neural correlate of consciousness (NCC)."
"In formal ontology, a branch of metaphysics, and in ontological computer science, mereotopology is a first-order theory, embodying mereological and topological concepts, of the relations among wholes, parts, parts of parts, and the boundaries between parts."
"To summarize: the moment we take seriously the fact that all measurement is a form of abstraction, we find ourselves compelled to question the logical possibility of measurement in the context of general relativity. Nowhere in that theory can we find the projective and mereotopological relations necessary to give meaning to measurement. On the other hand, bimetric theories become attractive both for their logical characteristics and their potential empirical consequences. Entire avenues are opened up for the possible reconciliation of macro and micro physics. If we elect to do this in the context of a Whiteheadian theory of nature, extension and abstraction, then we will additionally connect these physical theories within a mereotopological framework that is itself intimately connected to research in the area of general spatial reasoning. If we do not choose to do this in a Whiteheadian theory of nature, then we are obliged to say how the extensive deliverances of concrete experience link up with our abstractive processes of measurement to tell us what this “other” nature might be, and how this “other” nature can nevertheless be “really Real” despite the fact that our only access to it is through an abstract process of our own making. It should surprise no one to learn that I prefer the Whiteheadian solution."
Objects thus become syntactic operators, and events become intersections of nomological syntax in the common value of an observable state parameter, position.
The circle corresponding to the new event represents an attribute consisting of all associated nomological relationships appropriate to the nature of the interaction including conserved aggregates, and forms a pointwise (statewise) “syntactic covering” for all subsequent potentials.
Since this potential can only be specifically realized through the infocognitive binding of telesis, and localized telic binding is freely and independently effected by localized, mutually decoherent telic operators, deviations from perfect complementarity are ubiquitous. SCSPL evolution, which can be viewed as an attempt to help this complementarity emerge from its potential status in MU, incorporates a global (syntactic) invariant that works to minimize the total deviation from perfect complementarity of syntax and state as syntactic operators freely and independently bind telesis.
SCSPL incorporates the concepts of syntactic stratification and syntactic distribution. For example, because the laws of mathematics everywhere apply with respect to the laws of physics, the former distribute over the latter in the syntactic sense. Thus, where the laws of mathematics and physics are denoted by S1=LMS and S2 respectively, S1 distributes over S2, i.e. forms a syntactic covering for S2.
Essentially, this means that the laws of mathematics are everywhere a required syntactic component of the language of physics. With S2 is associated an SCSPL “sublanguage” called LO (with a letter O subscript). LO constitutes the world of perception, the classical objective universe of sense data traditionally studied by science. LO is contained in the telic-recursive, pre-informational phase of SCSPL, LS, which encompasses the cross-refinement of LO syntax and LO content from the pre-infocognitive aspect of SCSPL. The part of SCSPL grammar confined to LO incorporates certain restrictions to which LS is not subject; e.g., the grammatical portion of LO (S2) is fixed, distributed and supposedly continuous, while that of LS can also be mutable, local and discrete…in a word, telic.
Γ grammar is the generative grammar of SCSPL = (LS⊃LO). Γ grammar is unlike an ordinary grammar in that its processors, products and productions coincide and are mutually formed by telic recursion. Syntax and state, loosely analogous to form and content (or productions and products), are mutually refined from telesis through telic recursion by infocognitive processors. Production rules include the Telic Principle, distributed elements of syntax formed in the primary phase of telic recursion, and more or less polymorphic telons formed by agent-level telors. The corresponding modes of production are global telic recursion, informational recursion by distributed syntax, and local telic recursion.
The “words” produced by Γ grammar are not strings of symbols, but LO spatial relationships among parallel processors that can read and write to each other’s states. In effect, the states of its processors are roughly analogous to the symbols and strings of an ordinary language. The processors of Γ grammar thus function not only as transducers but as symbolic placeholders for observables and values, while their external states correspond to products and their state transitions realize the productions of the grammar. In other words, the states and state transitions of the processors of Γ grammar comprise a representation of Γ grammar, rendering SCSPL a dynamic self-modeling language or “interactive self-simulation”.
Γ grammar generates SCSPL according to the utility of its sentient processors, including the self-utility of Γ and the utility of its LO relations to telors in A. Γ and A generate telons on the global and local level respectively; thus, they must be capable of recognizing and maximizing the selection parameter υ (in the case of human telors, for example, this requires the QPS (qualio-perceptual syntax) and ETS (emo-telic syntax) components of the HCS (Human Cognitive-Perceptual Syntax)).
As such, they are responsible for telic recursion and may be regarded as the “generators” of Γ grammar, while the set Q of elementary physical objects are freely and competitively acquired by telons and thus occupy an ontologically secondary position.
Γ grammar is conspansive. Non-global processors alternate between the generation and selective actualization of possible productions, and thus between the generative and selective (inner expansive and requantizative) phases of conspansion.
The selective phase of an operator coincides with interactive mutual-acquisition events, while the generative phase coincides with the generation and selective actualization of possible productions through hological multiplexing. In conjunction with extended spatiotemporal superposition, conspansion provides the means of local (telic and informational) recursion.
It is instructive to experiment with the various constructions that may be placed on LS and LO. For example, one can think of LS as “L-sim”, reflecting its self-simulative, telic-recursive aspect, and of LO as “L-out”, the output of this self-simulation. One can associate LO with observable states and distributed-deterministic state-transition syntax, and LS with the metasyntactic Telic Principle.
One can even think of LS and LO as respectively internal and (partially) external to SCSPL syntactic operators, and thus as loosely correspondent to the subjective and objective aspects of reality. Where LS and LO are associated with the coherent inner expansion and decoherent requantization phases of conspansion, so then are subjective and objective reality, simulation and output, “wave and particle”.
In other words, the subjective-objective distinction, along with complementarity, can be viewed as functions of conspansive duality.
Where space and time correspond to information and generalized cognition respectively, and where information and cognition are logically entwined in infocognitive SCSPL syntactic operators intersecting in states and state-transition events, space and time are entwined in a conspansive event-lattice connected by syntax and evolving through mutual absorption events among syntactic operators, symmetric instances of generalized observation influenced by telic recursion.
Thus, time is not “fed into” the explanation of existence, but is a function of conspansive, telic-recursive SCSPL grammar.
To see how information can be beneficially reduced when all but information is uninformative by definition, one need merely recognize that information is not a stand-alone proposition; it is never found apart from syntax. Indeed, it is only by syntactic acquisition that anything is ever “found” at all.
That which is recognizable only as syntactic content requires syntactic containment, becoming meaningful only as acquired by a syntactic operator able to sustain its relational structure; without attributive transduction, a bit of information has nothing to quantify.
This implies that information can be generalized in terms of “what it has in common with syntax”, namely the syndiffeonic relationship between information and syntax."