Information Technology Reference
In-Depth Information
C
1
is unit literal, that is when C
1
= {L
1
},
}
ȶ
1
)
ȶ
2
-1
(9.6)
ȶ
2
-1
is inverse substitution, which is the only inverse substitution
ȶ
-1
satisfied with
t
ȶȶ
-1
= t given term and literal t and substitution
ȶ
. What is more, if
ȶ
=
{v
1
/t
1
,…,v
n
/t
n
},then:
ȶ
-1
= {(t
1
,{P
1,1
,…P
1,m1
})/v
1
,…,(t
n
,{P
n,1
,…,P
n,mn
}/v
n
) (9.7)
Here, P
i
,
mj
is the position of variable v
i
in t. Inverting resolution substitute
all t
i
on the position of of with v
i
{P
i
,
1
,…,P
i
,
mi
} in t.
EBL aims to construct an explanation tree with regard to training examples
from the initial description of goal concept by taking advantages of domain
theory (stored in the knowledge base). The whole process is generally based on
goal-driven inference. When the learning fails caused by explanation can not
continue lack of some particular rule, inverting resolution is used to overcome
this problem(Haibo Ma, 1990).
Domain theory (knowledge base) is represented by production rule. Some
pre-processing needs to be done to knowledge base to create a dependent tree
describing the relationship of rules. Therefore, domain theory is made up of a
group of generation rules and a dependent table.
The dependent table is the denotation of relationships among predicates based
on rules of the knowledge base. The table includes rules and semantic
information about predicates. A simple example is given to show the structure of
dependent table.
Knowledge base
Rule 1.Sentence (S
0
, S):- noun-phrase (S
0
, S
1
),
verb-phrase (S
1
,S).
Rule 2.noun-phrase (S
0
, S):-determiner (S
0
, S
1
),
noun (S
1
, S).
Rule 3.noun-phrase (S
0
, S):-name (S
0
, S).
Rule 4.verb-phrase (S
0
, S):-intransitive-verb (S
0
, S).
The dependent table is as following:
C
2
= (C
∪
{
L
1
Table 9.2
Predicate symbol
Head
Body
Basic predicate
sentence
1
Intransitive-verb, name,
determiner, noun
noun-phrase
2,3
1
Name, determiner, noun
verb-phrase
4
1
Intransitive-verb
