search

ChildMindInstitute / mhdb-tables2turtles

Text processing code to convert specific spreadsheets to RDF as initial content for the Mental Health Database (MHDB)
Other
0 stars 0 forks source link

wearables / apps / technology #17

Closed shnizzedy closed 6 years ago

shnizzedy commented 7 years ago

nail down the format for entering wearables, apps, technology in our database.

shnizzedy commented 7 years ago

First draft

shnizzedy commented 7 years ago

Second draft

shnizzedy commented 7 years ago

Third draft

On the first sheet, I color coded my opinions as follows:

shnizzedy commented 7 years ago

Fourth draft

We can't usefully encode everything in a single sheet unless we severely limit our scope. I tried to get the number of sheets down to as few as possible while still being useful. @anirudh4792, see if you can fit all 10 examples (or just the most difficult of the 10 examples) in this format and see if we're missing anything or if anything is unclear.

I compiled our collected high-level definitions. This is the kind of thing we could use a SPARQL query for.

anirudh4792 commented 7 years ago

  1. Not able to access entire worksheet (says 'its working') Assigned to @binarybottle for review of the 10 new examples considering Jon's comments

  2. "We can't usefully encode everything in a single sheet unless we severely limit our scope." This is not necessary or the rule of thumb at all. I just try to make it fit for my readability/for this particular set of examples as otherwise the reader will have to move about different sheets figuring out the numbers. Until possible and practical, I would like to keep things in a single sheet but will be happy to branch out when needed

  3. @shnizzedy the examples in the third draft are from the ten examples and figuring out the subjective state for each as they are sometimes the measure was quite tricky. Apart from that, things seem to fit

  4. Thank you for the definitions. It is clearer to me on what is expected.

binarybottle commented 6 years ago

After discussion with @anirudh4792, I think our initial four dimensions are reasonable:

  1. frequency
  2. duration
  3. intensity
  4. (cap)ability

I think we should also consider adding the following two:

  1. consistency (random to rigid)
  2. specificity (uniform to narrow context)

Example for 5: We never know when he'll show up for work. -> He shows up for work even on holidays. Example for 6: He always shows up for work at CMI but randomly shows up for work at NKI.

I think we can introduce a new behavior that will accommodate the following two, and so we probably don't need these: