Saturday, May 28, 2005
Notational Standards
discussion à
The Roadmap for Semantic
technology adoption à
The next generation knowledge tools
(application to medical science)
The concept that I have thought about, for years, is one in which information derived about gene expression function-structure relationships are encoded into structural taxonomy of technical terms and some simulation tools. For example: colored Petri nets might be used to assist researchers in communicating about what these terms mean.
The taxonomy terms would also be used to touch into a medical research literature as a means to provide very focused ontological modeling based information retrieval into research literatures that have NOT been indexed with metadata. The semantic extraction process would allow (just-in-time) fine-tuning of ontological models without having to have the text indexed manually with metadata.
(Note to Richard Ballard –
[101]) à
Gene expression "events" are identified as co-occurrence patterns using the SLIP methods, as well as stochastic/algebraic Latent Semantic Analysis, applied to numerical measurement of rates of growth, ratios between protein concentrations, and other standard measured
Key words: Simulation of complex metabolic pathways, knowledge management, structure function analysis, ontological modeling, data visualization.
We would use the "Actionable Intelligence Process Model"
http://www.bcngroup.org/procurementModel/to-be/actionable.htm
as an event loop. The key is to acquire scientific data purely and then bring this data into an environment where so called “semantic extraction” software provides an n-gram, generalized n-gram [1] , based measurement of the numerical data. The measurement process itself is isolated, or separated from the analysis methods, and variations on standard measurements are made available to the set of semantic extraction tools.
The measurement of gene expression would be made using some standard methods (Ben Goertzel might be able to help here) as well as some new methods. The output of the measurement would be encoded as co-occurrence patters - to reflect "unlabeled" emergent organization of the gene expression. A community, using knowledge management software, would work to label the co-occurrence patterns.
Labeling would occur as a function of terminological reconciliation processes between various members of a research team.
Real natural world ontology expresses as phenomenon. The expression involves a response degeneracy, or singularity, in function to structure determination [2] [3] [4] . Some work on this subject has extended the spin glass work that I was able to participate in 1988 [5].
The facts about double articulation of meaning and word structure [6] are not ones that is commonly recognized by, for example, Semantic Web community. The shallowness in not recognizing response degeneracy, or singularity, in function to structure determination creates problems for our community when program managers in the federal government decide to proto-type ontology system to use to "understand" threats and risks.
However, the technology that is being using in semantic extraction “from text” exposes the very same issues that one sees in theoretical physics and in the theories of human consciousness (see McLennan’s work [7]. ) McLennan’s work has been influential in the BCNGroup circle since the Pribram conferences started in 1991.
Science is beginning to separate itself from the reductionism of the twentieth century ... with books like "A Different Universe, reinventing physics from the bottom down”, by Robert B Laughlin" (2005). In this work, the real universe is seen as organized into interaction levels, with emergence being one of the cross scale phenomenon having locality and conservation issues (as in Bell's inequality).
psp
[5]
Kowalski, J. ; Ansari, A. ; Prueitt, P. ; Dawes,
R. and Gross, G. (1988.) On Synchronization and Phase Locking in Strongly
Coupled Systems of Planar Rotators. Complex Systems 2, 441-462.
Kowalski, J., Labert, G., Rhoades, B, & Gross,
G. (1992). Neuronal Networks With Spontaneous, Correlated Bursting Activity:
Theory and Simulations. Neural Networks, 5, 5, 805 - 822.