Let us assume that we deal with series of data items, some of which
may be contradictory. Moreover, different sources may give data a
different degree of trustworthiness which can make some pieces of
inconsistent data to be preferred. Lets us assume that p(A)
and q(A)
are contradictory and we receive, from source S1, p(A)
and, from
source S2, q(a)
. We may decide that: (i) p(A)
is true because S1 is
more realiable; (ii) or if S2 is more realiable, q(a)
is true, and any
value not a
(i.e., X = a) p(A)
is also true; (iii) or, if both
sources are equally reliable, them we have (at least) two different
models: one where q(a)
is true and another where p(A)
is true (also
for X=a).
scasp stream_data_reasoning.pl