Tuesday, January 17, 2006
[350] ß parallel discussion in “national debate” bead thread
This is part of a discussion
that will be moved to a Wiki page soon.
Ontological modeling with
category theory and convolutions
Post OWL strategy
[347] ß Judith Rosen’s recent communication
Communication to a working group;
Perhaps someone (each person) with actually current working experience with signal pathway phenomenon, might enumerate the list of 3 - 12 primary issues that biological researchers would like to be able to address, illustrate or research.
This listing should not be constrained by what might or might not be done using computer analysis of data from advanced instrumentation.
Descriptive Enumeration is often very easy for an domain expert to do... if he/she is in the right mood. Takes a few mins, even though the knowledge outlined may have taken decades to acquire.
Note that this is like the formation of a set of postulates and axioms (as historically seen in reference to Euclidean geometry - and which is only weakly similar to the OWL type "expressive logics"). The key notion is completeness (in some sense) and independence of each descriptive phrase (or concept). Note that, unlike the OWL "axioms" the DE axioms are a set of concepts related to a domain of inquiry.
The notion of consistency is wonderfully different from "completeness" and "independence"... while being part of the classical notion of a "formal system". The concepts related to inference; ie abduction, induction and deduction; can be moved aside - to be added later when/if appropriate.
So, an ontological cover (over a domain of discourse) is a set of concepts (without any machinery for "inference). The retrieval of properties, attributes and relationships defined within the set of concept representations is not a matter of theorem proving, but merely one of "retrieval".
I developed a methodology that is now used in some graduate knowledge management courses...
The first description of DE was in a e-GOV conference in 2000.
http://www.bcngroup.org/area3/pprueitt/private/KM_files/frame.htm
slide six.
Note that there is a "community based" knowledge elicitation process. Knowledge management methodology based on this process asks many domain experts to make descriptively enumerated lists independently; and then to form, 1) viewpoints, and 2) community consensus.
A second step might be to talk about available physical instrumentation for acquiring primary data. This discussion should use easy to understand language that a program manager at NSF might understand. (In response to Joanne note about work that needs to be done and needs to be funded, so that we can have the dedicated time to do this work.)
A tutorial might then be developed to expose what the metadata and categorization (by concept ontologies, also sometimes called semi-formal ontology) issues are in light of these primary research issues. This tutorial can be published at conferences and can be the basis for group proposal development.
The direct result of this work would be to clarify the mission of BioPAX, while staying focused on biological science. The indirect result would be a group based resource (like a wiki) that could be used to evangelize the work done on cell and gene expression.
I am now funded specifically to develop semantic interpretation environments that use the OASIS service oriented architecture and semantic execution environments to produce portal based inter-community collaboration infrastructure. So there are some post OWL tools that will become available... given that one has a well-defined OWL front end to a information base.
The metaphor between the “natural complexity” of cell and gene expression, and the natural complexity of social and business expression allows me to justify working on “next step” (Post OWL) Semantic Interpretation Environment (SIE) as a possible new OASIS standard (being considered internally by my group).
Dr Paul Prueitt PhD (mathematics and quantum cognitive neuroscience)
The Taos Research Institute
Taos New Mexico