Information Technology Reference
In-Depth Information
Bel X
0,7
0,5
0,4
Ability Y
0,2
0,4
0,8
X
Figure 3.2
Relationships among different evalutative Beliefs and their strenghts
modest ability of Yody (0.40) and at the same time Xania believes (with high certainty (0.70))
Yody belongs to a category which, generally (for its own features), performs very badly (0.20)
on those kinds of task.
At this stage of the analysis it is also possible to have concurrent hypotheses with similar
values (the sum should not necessarily be equal to 1 : again, they are not coherently integrated
but just singularly evaluated). The meta-belief (we can also consider it the strength of this
belief) on Y 's ability has to consider the different beliefs about this quality 6 and, on the basis
of a set of constraints (see later), define the degree of credibility DoC of the specific quality.
In general, we could write that:
DoC X ( Qual - i ( s 1 ,... sn ) , Y (
τ
))
=
F X , Y ( Bel X ( Str 1 Qual - i s 1 Y (
τ
))
,
Bel X ( Str 2 Qual - i s 2 Y (
τ
))
,...,
Bel X ( Str n Qual - i snY (
τ
)))
(3.2)
In other words, the degree of credibility for X ( DoC X ) about the i -th Quality of Y on the task
τ
( Qual iY (
τ
) ) on the basis of the n belief sources (S 1 ,
...
,S n ) is a function (depending on X, Y
6 Here we are considering Y 's ability as a quality of Y . We have used the term quality in Chapter 2 at a more detailed
level: the set of features characterizing an entity. In fact Y 's ability might be considered as a meta-quality (composed
by a set of more specific qualities).
 
Search WWH ::




Custom Search