OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

soa-rm-ra message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Re: [soa-rm-ra] Resend with readable jpg: My takeway (while listening in to another telecom)

see inline

On Jun 26, 2009, at 5:14 PM, Estefan, Jeff A wrote:

Certainly agree that Trust is perception, but I do feel comfortable with couching Risk is as perception since Risk is exposure (rather than perception) to something undesirable happening.  That level of exposure (i.e., risk exposure) requires analysis.  And we probably don’t need to go this far in the RA, but it is typically measured in terms of likelihood and consequence.

I understand your point but many assessments of risk are done as informally as assessments of trust.  I would suggest the definition in terms of perception, i.e.

Risk is an actor's private perception that another actor's actions and other factors will result in undesirable real world effects.

but having text following the definition covering your points of risk often subject to formal analysis to quantify level of risk.  In reality, the definition of a risk threshold is often contentious and a perception of when who feels the risk is low enough, e.g. Yucca Mountain.  But I am open to further wordsmithing.

Ken, what do you mean by “and other factors” in your recent proposed definition of both concepts? 

See Rex's email to which I replied.  He introduced the "and other factors" and explained his rationale.  

I think we need something more solid or at least provide examples; otherwise, we should just drop those words.

I would be willing to drop it.  Rex?

Finally, what was wrong with the Trust definition in the latest RA draft that is tied to RWE?  “Trust is an actor’s private perception of the commitment another actor has to a goal together with an identifiable set of real world effects associated with that goal.”  Perhaps we need a reprieve on RWE from the RM just like we did with Willingness.  Recall we said RWE (or set/series of effects) is the result of an interaction and an interaction is “an act” as opposed to “an object.”  We also said the RWE are couched in terms of changes to shared state.

That's exactly it -- I reprieved the RWE connection, leading to

Trust is an actor's private perception that another actor's actions and other factors will result in desired or neutral real world effects.

The other actor's goals are really not the issue but rather the RWE that we want to see.  For example, in Virginia, a common goal is to have children show they are educated in history.  The official RWE desired is for them to pass a test asking them names and dates.  I find that totally inadequate as a desired RWE and thus my trust in the state's ability to educate my children in history is far less than would be reflected in our shared goals.

Just trying to get some grounding here.  Seeing so many different proposals on how to model Trust these days, it’s getting quite confusing while fully recognizing it’s a very difficult concept to model.

I believe there is a linear progression (with side discussions on UML modeling) in the interchange between Rex and me.  A branch was Frank's proposal (split in Evidence) and some comments of preference for the other one (single Evidence).

 - J

Ken Laskey
MITRE Corporation, M/S H305      phone: 703-983-7934
7515 Colshire Drive                         fax:       703-983-1379
McLean VA 22102-7508

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]