J. Dorre and A. Eisele. Determining consistency of feature terms with distributed disjunctions. In D. Metzing, editor, GWAI-89, 13th German Workshop on Artificial Intelligence, pages 270--279. Informatik Fachberichte 216, Springer-Verlag, 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....disjunctive normal form, which will usually cause an exponential blow up in size. Kasper [21, 20] and Dorre and Eisele [11] have proposed better unification algorithms for feature terms that try to avoid pushing up unions as much as possible. A new constraint solving method of Dorre and Eisele [9] introduces so called distributed disjunctions and works on feature constraints rather than feature terms. Kasper [22] investigates the use of feature terms with implications (an implication S T is equivalent to :S tT ) for modeling systems in systemic grammar. He outlines a unification method ....

J. Dorre and A. Eisele. Determining consistency of feature terms with distributed disjunctions. In D. Metzing, editor, GWAI-89, 13th German Workshop on Artificial Intelligence, pages 270--279. Informatik Fachberichte 216, Springer-Verlag, 1989.

....the FA A 1 which recognizes the language L(A 1 ) a b) c. X Y c a,b A 1 A 1 consists of the two states X and Y; therefore, we have to define two types X and Y , where Y (given in (5) is only an instantiation of a final configuration. Note that we make use of distributed disjunctions [5] (depicted by the disjunction name 1) in the definition of X to express the covariation between edges and successor states: if a is processed, use type X (and vice versa) if b is processed, use again type X , but if c is chosen, choose type Y . X j 2 4 non final configuration EDGE 1 (a b ....

Jochen Dorre and Andreas Eisele. Determining consistency of feature terms with distributed disjunctions. In Dieter Metzing, editor, Proceedings of 13th German Workshop on Artificial Intelligence, GWAI-89, pages 270--279, Berlin, 1989. Springer.

....term is inconsistent: top(l 1 ) f d1 ; f d2 ; t 1 gg l 2 ) f d2 t 2 ; f d1 t 3 ; gg) Therefore all inconsistent contexts must be stored globally. A detailed description of the ffi term calculus is given in [8] Similar systems using distributed disjunctions can be found in Dorre and Eisele [9], and in Maxwell and Kaplan [13] We see two advantages of our approach compared to Dorre and Eisele s: 1) Their formalism does not treat attribute disjunctions. 2) In our system the unification of disjunctions is defined more abstractly, in that the method of finding appropriate disjunction ....

Jochen Dorre and Andreas Eisele. Determining consistency of feature terms with distributed disjunctions. In D[ieter] Metzing, editor, Proc. of the 13 th German Workshop on Artificial Intelligence, volume 216 of Informatik Fachberichte, pages 270--279. Springer, Berlin, 1989.

....with such disjunctive feature terms is worst case intractable, nobody wants to give them up they are simply necessary for the particular kind of problem one wants to solve. The challenge then is to identify the structure of normal cases and to design algorithms that deal with these normal cases [12, 6] which are informally characterized by the fact that humans can deal with them effortlessly. Summarizing, after discovering that a representation formalism leads to intractable reasoning problems, it is worthwhile to analyze how the representation formalism is used. If it turns out, that the ....

J. Dorre and A. Eisele, Determining consistency of feature terms with distributed disjunctions, in: D. Metzing, ed., GWAI-89. 13th German Workshop on Artificial Intelligence, (Springer-Verlag, Berlin, West Germany, 1989) 270--279.

....concept expressions are polynomially bounded in size by the size of the terminology. Probably, it is not possible to find always such simple restrictions. However, it may turn out that for all cases occurring in practice, sophisticated algorithms can avoid an exponential explosion (see e.g. [12, 25]) 4.4 The Instance Determination Problem Viewed in isolation, the results reported in Table 1 and Theorem 3 are discouraging. However, as already pointed out above, Theorem 3 may not affect the efficiency of subsumption in practice, and, moreover, intractability of term subsumption may not be ....

J. Dorre and A. Eisele. Determining consistency of feature terms with distributed disjunctions. In GWAI-89. 13th German Workshop on Artificial Intelligence, pages 270--279, Springer-Verlag, Berlin, West Germany, 1989.

....completing the merging of t 1 and t 0 1 in the context fhd 1 ; 1i; hd 2 ; 1i; hd 3 ; 1ig. Therefore all inconsistent contexts must be stored globally. A detailed description of the ffi term calculus is given in [6] Similar systems using distributed disjunctions can be found in Dorre and Eisele [7] and Maxwell and Kaplan [11] We see two advantages of our approach compared to Dorre and Eisele s: 1) Their formalism does not treat attribute disjunctions. 2) In our system the unification of disjunctions is defined more abstractly, in that the method of finding appropriate disjunction ....

Jochen Dorre and Andreas Eisele. Determining consistency of feature terms with distributed disjunctions. In D[ieter] Metzing, editor, Proc. of the 13 th German Workshop on Artificial Intelligence, volume 216 of Informatik Fachberichte, pages 270--279. Springer, Berlin, 1989.

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