zhengj2007
commented
9 years ago From steschu@gmail.com on June 24, 2013 22:26:25
See an updated version of the slides at https://bfo.googlecode.com/svn/trunk/docs/BFO2_Buffalo20130601.pptx Open issues with TQC:
- Identity of TQCs
- Participation of TQCs in processes
From steschu@gmail.com on May 27, 2013 16:57:02
This is a short update of the TQC approach during the BFO2 meeting in Buffalo and some iterations afterwards, done by Janna Hastings, Stefan Schulz, and Fabian Neuhaus, getting very useful feedback by Alan Ruttenberg and Barry Smith.
This is work in progress, and it should not be recommended (just as other, competing approached) for community adoption until all the technical details of these approaches have been fully worked out and evaluated. The existing agreement to release BFO OWL with an official subset that exludes the temporalized relations (and corresponding adoption of atemporal relations in RO for the meantime) is consistent with this requirement.
The main idea of the TQC approach is that it allows a simpler relation hierarchy (no TRs) and has as default (simplest) case permanent generic relatedness. However, this is at the expense of requiring a proliferation of instances (as instantiation is necessarily time-region-linked). There are still open questions as to how the approach would work out for all the different types of relation and all the properties that relations might have in OWL.
Next steps, as agreed with Alan, will that we will work together, within a short time frame, to apply both approaches (Alan's TRs and the new TQCs) to a defined set of core use cases (instantiation of non-rigid classes, permanent generic parthood, and so on) and thereby evaluate the approaches side-by-side with regard to important community considerations such as usability, complexity, formal correctness, ease of migration for existing ontologies, and importantly time to perform reasoning over large ontologies with actual OWL implementations.
The output of this process (within two months) will be a substantially revised version of what is now the TQC paper draft ( https://bfo.googlecode.com/svn/trunk/docs/TQC/TQC-current.pdf ) that not only presents one of the approaches but instead thoroughly compares and evaluates both approaches side by side. This would then serve as a basis for community participation in the decision process as to which method BFO OWL "officially" adopts - or indeed, whether we sanction both and allow users to choose which they adopt.
The current state of the TQC approach is summarized in this slideset: https://bfo.googlecode.com/svn/trunk/docs/Temporally_Qualified_Continuants_BFO2_Buffalo20130521.pptx paralleled by an example owl file https://bfo.googlecode.com/svn/trunk/src/ontology/owl-schulz/tqc6.owl Comments and suggestions are welcome!
Original issue: http://code.google.com/p/bfo/issues/detail?id=179