W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > June 2007

Re: Evidence

From: Adrian Walker <adriandwalker@gmail.com>
Date: Mon, 11 Jun 2007 14:45:16 -0400
Message-ID: <1e89d6a40706111145l455af6abg55e5b849c1539487@mail.gmail.com>
To: "Matt Williams" <matthew.williams@cancer.org.uk>
Cc: public-semweb-lifesci@w3.org
Hi Again Matt --

You wrote...

It is also not "definitely necessary to relate hypotheses to evidence
with probability" (although it may be useful). There are a load of other
techniques that don't use probability: e.g. Wigmore Charts (from 1930's
onwards) and more recently, non-monotonic logical techniques.

Yes indeed, non-mon -- as in the much maligned negation-as-failure-to-prove
-- should be useful in many circumstances.

Another technique is to document your real world confidence in each
reasoning step, by writing your inference rules in English in the first
place.  For example a conclusion might be

  "at 3:31 pm on Monday the available facts facts appear to show that the
patient's condition is no worse"

The evidence for this can be a chain of reasoning steps, in English, tracing
back to the facts that were used, and also to missing facts (e.g. patient
temperature above 101 degrees) whose absence was used.

As you know, there's technology online at the site below that supports this
approach, along with other more classical approaches.

Apologies to folks who have seen this before.

                                                -- Adrian

Internet Business Logic (R)
A Wiki for Executable Open Vocabulary English
Online at www.reengineeringllc.com    Shared use is free

Adrian Walker
Reengineering

On 6/11/07, Matt Williams <matthew.williams@cancer.org.uk> wrote:
>
>
> I changed the subject line to make it more specific.
>
> I think that Evidence is a tricky, slippery subject. It seems to be both
> traces (i.e. records of something) and in many cases, inferences. Those
> inferences probably shouldn't be called evidence, but they are the
> reason that some data are considered evidence, and others not, and hence
> often get included.
>
> To take the radiology example below
>
> >> So evidence is a function of the facts, the
> >> analysis method, the method of inference, and perhaps even the
> >> observer (e.g., if the evidence is a radiology image or physical
> >> exam, there is inter-observer variation).
> >> And it's definitely necessary to relate the hypotheses to the
> >> evidence with probabilities
> >>
> I would suggest that the interpretation of the evidence is a function of
> the facts (plus other things). However, the facts are not stable (e.g.
> with a physical examination) and may conflict with each other; therefore
> inconsistency is not a just a matter of which inference procedure you
> choose, it is also a matter of which facts (your premises) you start from.
>
> It is also not "definitely necessary to relate hypotheses to evidence
> with probability" (although it may be useful). There are a load of other
> techniques that don't use probability: e.g. Wigmore Charts (from 1930's
> onwards) and more recently, non-monotonic logical techniques.
>
> For a good intro. I would recommend David's Schum's book "The Evidential
> Foundations of Probabilistic Reasoning". Also,a look at the evidence
> science website might be good: http://www.evidencescience.org/
>
> HTH,
>
> Matt
>
>
Received on Monday, 11 June 2007 18:45:21 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:00:48 GMT