Main Navigation

User login

Devices Ontology Working Group Meeting 2008.01.22

Agenda for the 2008.01.22 (1600 GMT) meeting of the Devices Ontology working group. (Minutes follow the agenda.)

A. This and That

Note Taker.
Comments/corrections of last minutes.
Comments/corrections on Agenda.
Roll Call.

B. Presentation: Dr. Femke Reitsma, Edinburgh University, COMPASS

(Pending availability.)

COMPASS is an organization working on an ontologically driven knowledge infrastructure for the marine domain. One of the areas they will be considering Very Soon is instrument ontologies. I've invited them to talk a bit about their project, and to hear us talk about our project, to get an idea of what synergies may result.

C. Going over Roy Lowry's "P012" vocabulary product

Roy Lowry had some success producing a candidate parameters vocabulary by synthesizing information from his existing database. (This vocabulary's content seems to align well with the type of information in the CF vocabulary.) He previously circulated the resulting Excel document in email to the list (on 2007.12.17T161200Z). He would like feedback from the group on this list as soon as possible. Questions to answer:

Is this more like what we need than the full-blown monster?
If so, what's missing or wrong with it?

With some guidance, Roy can try to build a complete list from the underpinning semantic model using automated procedures.

D. Review of Use Cases, Particularly 11 and 12

Nan Galbraith observed that at least one use case was missing, and added it while we were trying to pull together the meeting. It is now listed as Use Case 11.

As I looked at these and other use cases, I wanted to suggest/make a few changes. Since we are starting to have a history of changes, I first moved the use cases text into the subversion repository, and out of the use case page. Then I made the changes to the use cases, so you can see what I changed.

You can browse the subversion repository (most interesting files are in the docs/trunk directory at the moment), see individual use cases linked at the use cases page, or browse each use case via the repository.

I apologize for the clunky format, and the fact we can't easily see all the use cases on one page now. I am open to being helped.... The repository svn URL is svn://mmi.mbari.org/ontdev; see me for passwords or help, ideally after 25 January.

E. Status of Action Items

If we have time we will go over the Action Items.

See the TRAC page for the latest list of action items (due by current telecon) and list of action items from most recent telecon.

This and That

No corrections to minutes.

P012 Vocabulary

Presentation

Roy presented the work he has done on creating a potential set of parameter terms from his BODC vocabulary. In the original monster database there were biological and non-biological terms, but he worked with only the latter category. By deleting elements aside from the actual measurement being performed (like how the measurement was performed), Roy has shown a process that could create a comprehensive list of candidate 'observation parameters' for us to use.

The questions clarified several points, among them:

These terms will describe parameters measured by an instrument, not those used to describe a data set.
Duplicates need to be eliminated.
We will not include method of measurement.
Parameters that include units are not desirable, but may be hard to systematically remove. We should look at this when the whole vocabulary is generated, see if these cases are the exception or closer to the rule. John suggested such parameters should only be there if that is the actual measurement being made. [But much later, upon getting the point finally, he decided that even if it is the actual measurement, units should not be incorporated. Although ratios are incorporated into CF terms....]

All thought the vocabulary process looked viable. We proposed the following steps:

Roy produces a vocabulary using all the terms.
That vocabulary is compared with CF (likely very little overlap).
That vocabulary is compared with MBARI's observation variables.
Any result of all of that (possibly including merged terms) is made available for all to use to describe their instruments.

Review of Use Cases

John had to do a last-minute patch so use cases could be seen on-line -- port 8081, the SVN browsing port, is not available from all sites. It will be really nice to automate this patch someday.

Use case 11 proved acceptable to everyone.

There was some confusion about use case 12, the end conclusion of which was that there may be 2 use cases there. John is going to rewrite and show to Luis for comment/further clarification.

Possible Synergies

Jianping Mao, the Technical Lead of GCMD, was introduced at today's meeting. John described the degree of similarity between the work GCMD needs to do to make a user-friendly instruments keyword list, and the categories and activities being pursued by the Device Ontology project. Hopefully the two projects can find useful ways to collaborate.