Laurence Cholvy. Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle, 4(1), 1990.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Skolem functions with universally quantified variables is an important component of the generalized characterization of answer we develop in the context of resolution theorem provers. Cholvy and Demolombe have used a logic framework for investigating generic answers [Cholvy and Demolombe, 1986, Cholvy, 1990] They define an intensional answer to a query Q(X) as a set of formulas A(X) such that 8X(A(X) Q(X) follows from the knowledge base. They use first order predicate calculus for knowledge representation and resolution with the set of support strategy for reasoning. The resolvents generated in ....

.... 285) 11. f(ANSWER (LIKESTOEAT PUSS 523) TUNA 523) TUNA 523) g ) LIKESTOEAT PUSS 523) It is generally accepted that such rules should be written as closed sentences, where all of the variables are universally quantified [Cholvy and Demolombe, 1986, Imielinski, 1987, Motro, 1989, Cholvy, 1990] Universal quantification is appropriate, because all variables in clauses are implicitly universally quantified. The final forms for clauses 10 and 11 are shown in Table 4.7. Table 4.7: Rules as Answers Clause Answer 10. 8 285 [ FISH 285) LIKESTOEAT PUSS 285) 11. 8 523 [ TUNA 523) ....

[Article contains additional citation context not shown here]

Laurence Cholvy. Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle, 4(1), 1990.

....2.2 Generic Answers A generic or intensional answer is generally associated with a rule that expresses general properties of classes or groups of objects that satisfactorily answer a question. Generic answers have been examined by researchers working in rule bases [Cholvy and Demolombe, 1986; Cholvy, 1990] as well as databases [Reiter, 1978a; Reiter, 1986; Imielinski, 1987; Motro, 1989] and cooperative answering [Lehnert, 1978; Kaplan, 1981; Webber, 1986] Our definition of generic answers is based on variable sharing and requires the following definitions: Variable Sharing Set of a Literal. The ....

Laurence Cholvy. Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle, 4(1), 1990.

.... Ifcompletion is e#ecteddurin theman agemen t ofpatien ts on adesignT plan then the clinO( T (unO( ely to be the sameinT)I7F4T as theplan design7T is committed to those choices made by the plan designT at the time ofun 77 TB) On possible solution is to restrict untrictT to in varian t solution [1], i.e. those based ondon the in tenO#I(T part of the database, without use of facts in theexten(TB) part. However, on would nn to weaken thisrestriction to excludeonl medical facts, but allow solution based on n) facts in troduced by thedesign4 of the plan In future work weplan tocon7(II more ....

Cholvy L. Answering queries addressed to a rule base.I n: Revue d'intelligence artificielle, 4(1), pp 79--98, 1990.

....no answers will be produced. This is the approach taken by Prolog and resolution theorem provers. The type of answer produced using this approach is termed extensional, or specific, and the form of such an answer can be characterized as a set of a1. an such that Q(a1. an) Cholvy and Demolombe [2, 3] expanded upon Green s work by looking at resolution in a rule base with no ground terms. In this situation, the empty clause is never produced, yet answers in the form of rules rather than facts are discovered. Such answers are termed intensional or generic. The general form of these answers is: ....

Laurence Cholvy. Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle, 4(1), 1990.

....An extensional answer to this question might be the fact Mary likes John. Prolog employs resolution in this manner, answering questions with facts, and failing to answer questions when no relevant facts are contained in the rule base being searched. Cholvy and Demolombe (Cholvy Demolombe 1986; Cholvy 1990) expanded upon Green s work, using a resolution theorem prover to find intensional answers. An example of an intensional answer to the above question is Mary likes boys. The reason this is considered a rule is that boys is a category rather than an individual. Thus, the answer given above might ....

Cholvy, L. 1990. Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle 4(1).

....is warranted when the intensional answer is simpler than the query itself, or explicates something about the query and database. Intensional answers were introduced by [56] in which IAs are viewed as rewrites of a query which preserves its semantics (and, thus, is equivalent to the query) In [16] and [17] IAs are considered under the sufficiency criterion (that is, the IA is subsumed by the query) While intensional answers can be considered without integrity constraints, ICs allow more IAs to exist: the query and IA are logically equivalent with respect to the ICs and the database. In ....

L. Cholvy. Answering queries addressed to a rule base. Revue d'intelligence artificielle, 4(1):79--98, 1990.

....IA: Q (A) prerequisite (A, B) C 2 ) Q ( MATH 400 ) C 3 ) The IA in response to the query Q 6 is composed of the clause C 2 and the fact C 3 . If C 2 and C 3 were added to the database and the query h Q (A) i asked, the same answers would result as do with query Q 6 . Cholvy and Demolombe [8, 9] consider IAs under the sufficiency criterion (II) An IA represents sufficient conditions for a query to be true if all ground substitutions of the IA are correct (ground) answers to the query. They focus on IAs that are invariant. An IA is invariant if it implies the query by the sufficiency ....

L. Cholvy. Answering queries addressed to a rule base. Revue d'intelligence artificielle, 4(1):79--98, 1990.

....when the intensional answer is simpler than the query itself, or explicates something about the query and database. Intensional answers were introduced by [Imieli nski, 1988] Imieli nski viewed IAs as a rewrite of the query which preserves its semantics (and, thus, is equivalent to the query) In [Cholvy, 1990] and [Cholvy and Demolombe, 1987] IAs are considered under the sufficiency criterion (that is, the IA is subsumed by the query) While intensional answers can be considered without integrity constraints, 25 It is important to note that sometimes, due to the closed world assumption, different ....

Cholvy, L. (1990). Answering queries addressed to a rule base. Revue d'intelligence artificielle, 4(1):79--98.

....scheduled to land after 11 p.m. can be computed either from the flights registered in the information system, or from the above mentioned prohibition of late (scheduled) landings in Zurich. Researchers have investigated so called semantic query evaluation methods e.g. Chakravarthy et al. 1990, Cholvy 1990, Gal and Minker 1988) that, if possible, make benefit of (satisfied) integrity constraints for answering queries like this one. Although yet not used in marketed database systems, these techniques are promising. The framework of the present paper does not preclude them. In the following, an ....

Cholvy, L. (1990) Answering Queries Addressed to a Rule Base. Revue d'Intelligence Artificielle, 1, 1, 79--98.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC