proposal for gathering community knowledge

Dear all,

one thing that is always a concern in any community is how to retain knowledge and ideas generated by debate and discussion. I have spent some years in the community informatics arena, and this is a well-known problem.

What it comes down to is this. Interesting, sometimes long discussions occur about topics like "what is an episode", "age", "confidentiality" and so on. Many good points are made, and there may even be a consensus view reached due to the discussion. Then a new discussion occurs. What is lacking is a summary/précis of the material - a synthesis if you like. I suspect that most discussions which generate 50 pages of email could be summarised in about 2 A4 pages. Now, there are various technological attempts to help with this sort of process, such as Wiki and other online document modification systems. However, I have never been convinced that the best way still isn't to get a good human brain on the job, to produce a summary the old-fashioned way - by reading and thinking! The kind of document we are looking for is something ranging from an FAQ entry, to up to say 5 pages of detailed description; it would then be added to the online knowledge pages of the community.

My proposal is therefore that for each substantive discussion on a topic of interest, that an editor be found to produce a simple textual summary of the problem, arguments, and conclusions (or lack thereof; and reasons for that) which could then be incorporated into the openEHR website. Appropriate attribution for the editor and discussants would of course be included. This growing knowledge-bank would become an invaluable resource. Another kind of information we often get from clinicians (and technical people) is "use-cases" or "scenarios" - a description of how some process really works in their experience.

Note that there might not be a 1:1 relationship between a mailing list thread and a knowledge page - there are pearls of wisdom all over the place. However, a lot of raw material could I think could be turned into a decent 2-pager with a only couple of hours' work.

In concrete terms, here is how it could work:
- someone self-nominates as an editor of a topic for a knowledge or use-case page; it might be an expert in the subject, or a complete generalist who happens to be good at synthesising expert debate into expert knowledge
- they would generate a first draft, which would be put on a drafts section of the knowledge pages or use cases area of the website (doesn't exist yet)
- they would solicit comment and rework as necessary
- the final version would be posted in the relevant part of the website

Of course, nothing is ever "final" - the idea is to make later debate and thinking more efficient, as well as to capture domain thinking in a form that can be used by implementors. No heavy process would be imposed - anyone could submit a first draft and stop there if they wish. However, the editor would be encouraged to take the process through, from start to delivery; openEHR will just handle the details of putting the information on the website.

The first step I suggest would be to generate a list of possible topics for knowledge pages, use cases or other FAQ material - by going through the discussion archives, which are available online in thread- and date-indexed form. Perhaps 10 topics would do initially. Other sources of thought like hallway discussions and standards meetings arguments should also be considered.

Any takers?

- thomas beale

Thomas, just noted that this nice suggestion of your has not
evoked a response so far ... opening the batting here :wink:

What it comes down to is this. Interesting, sometimes long
discussions occur about topics like "what is an episode",
"age", "confidentiality" and so on. Many good points are made,
and there may even be a consensus view reached due to the
discussion. Then a new discussion occurs. What is lacking is a
summary/pr?cis of the material - a synthesis if you like. I
suspect that most discussions which generate 50 pages of email
could be summarised in about 2 A4 pages. Now, there are various
technological attempts to help with this sort of process, such
as Wiki and other online document modification systems.
However, I have never been convinced that the best way still
isn't to get a good human brain on the job, to produce a
summary the old-fashioned way - by reading and thinking! The
kind of document we are looking for is something ranging from
an FAQ entry, to up to say 5 pages of detailed description; it
would then be added to the online knowledge pages of the
community.

[rest snipped]

o The proposal of summarising important discussions is nice. It
  is also quite unique.

o Implementation would be a hurdle, because of the following :-

  a) You need the co-ordinating agency first, and then the core
     group needs to set the ball rolling ...

  b) Voluntary effort always works for intellctually stimulating
     things. IOW, the satisfaction to effort ratio would have to
     be high. In this type of thing, with smaller expected ratios
     voluntary jobs would have to be done military style: "One
     must be detailed to volunteer" :wink:

  c) Need a proper flow chart/ UML of the action plan :wink:

o I can help in the Medical portions. I am a medical graduate of
  1976 vintage, with a Diploma in Aerospace Medicine and MD.I am
  also an unqualified programmer/ IT man of sorts, with initial
  forays into programming since CP/M days, more or less self
  taught, having learnt tits and bits from works of 'gurus' like
  Knuth and Wirth. OOP is a oops situation for me but I am quite
  okay at scripting (shell, perl, php) and K&R C (programmed in
  pascal style) ! I am totally ANTI-M$, and on Linux/ *BSD since
  1995.

o If above QRs fit your bill, count me as an extra hand in for
  medical portions only ...

Cheers

Dr USM Bish
Bangalore

USM Bish wrote:

[rest snipped]

o The proposal of summarising important discussions is nice. It
is also quite unique.

o Implementation would be a hurdle, because of the following :-

a) You need the co-ordinating agency first, and then the core
    group needs to set the ball rolling ...

b) Voluntary effort always works for intellctually stimulating
    things. IOW, the satisfaction to effort ratio would have to
    be high. In this type of thing, with smaller expected ratios
    voluntary jobs would have to be done military style: "One
    must be detailed to volunteer" :wink:

I hope there would be satisfaction from having produced a nice readable article (let's call it that for now) which will be read by many others later on. Particularly if we index and present all the articles well.

c) Need a proper flow chart/ UML of the action plan :wink:

I think this is a good idea, although I often get flak for suggesting such things - too much process for some people.

o I can help in the Medical portions. I am a medical graduate of
1976 vintage, with a Diploma in Aerospace Medicine and MD.I am also an unqualified programmer/ IT man of sorts, with initial forays into programming since CP/M days, more or less self
taught, having learnt tits and bits from works of 'gurus' like Knuth and Wirth. OOP is a oops situation for me but I am quite
okay at scripting (shell, perl, php) and K&R C (programmed in
pascal style) ! I am totally ANTI-M$, and on Linux/ *BSD since
1995.

You are eminently qualified obviously. I am wondering if the best approach for this "editor" job is to get a sort of disinterested editor (who is reasonable at dissertation however) to collate existing discussion material into an article - a bit like a journalist or popular science writer - rather than an expert to write an article which might be more like an academic paper extolling one point of view, but perhaps not representing the discussion adequately. At least these two styles or writing are well-recognised.

Perhaps as an experiment, we can just use this current topic of "age". Nathan Lea at CHIME at UCL has offered to have a go at editing something - he is an excellent writer & researcher, but not a doctor - this will mean he will use pieces of what you and other clinical people have said in various posts (Sam has another nice summary a few articles earlier). My suggestion is to try an experiment - let's just see if we as a group can produce:
- an article summarising the "age" discussion. Needs decent headings to be thought up; can;t be too long; should cover all informed opinions found in posts
- a process or flowchart for doing this habitually

Reactions?

- thomas beale