K. von Luck, B. Nebel, C. Pelatson, and A. Schmiedel. The anatomy of the back system. Technical Report KIT- Report 41, Technische Universitat Berlin, Berlin, 1987.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....are widely used in many application areas, to date they have focused on representing structured conceptual descriptions, or concepts. A critique of contemporary TKR which argues for greater expressiveness is [ Doyle and Patil, 1991 ] One limitation of current terminological systems, e.g. BACK [ von Luck et al. 1987 ] CLASSIC [ Borgida et al. 1989 ] K Rep [ Mays et al. 1991b ] and LOOM [ MacGregor and Bates, 1987 ] is their inability to represent and reason with complex compositions of concepts, such as constraint networks where each node is described by a concept. Plans are central to many areas ....

K. von Luck, B. Nebel, C. Pelatson, and A. Schmiedel. The anatomy of the back system. Technical Report KIT- Report 41, Technische Universitat Berlin, Berlin, 1987.

....definitions alone. We call thesis (5) the restricted classification thesis. These theses were promulgated in the context of the descendants of kl one [8, 37] namely nikl [41, 26, 15] the New Implementation of KL one) kl two [41] kandor [30] Doyle Patil krypton [6] and back [42, 27]. While kl one itself was designed earlier, without concern for language restrictions, the designs of krypton, kandor, and back involved theoretical analyses (e.g. 19] and [27] to see what linguistic constructs could be efficiently classified and what constructs should be included or forbidden. ....

K. v. Luck, B. Nebel, C. Peltason, and A. Schmiedel. The anatomy of the back system. Technical report, Technische Universitat Berlin, Sekr. FR 5-8, Franklinstrasse 28/29, 1000 Berlin 10, Federal Republic of Germany, 1987.

....variables, feature symbols and set descriptions to ALC is investigated providing an alternative method for integrating concept languages and feature logics. It is shown that set descriptions can be translated into the so called number restrictions available within concept languages such as BACK [von Luck et al. 1987]. However, the propositionally complete languages ALV and ALS investigated in [Manandhar, 1993] are PSPACEhard languages which do not support set operations. The work described in this paper describes yet another unexplored dimension for concept languages that of a restricted concept language ....

K. von Luck, B. Nebel, C. Peltason, and A. Schmiedel. The Anatomy of the BACK System. KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, Germany, 1987.

....since users of a restricted DL knowledge base use ad hoc methods that jeopardize the soundness of the whole system ( 11] The LOOM system ( 26] takes a contrary approach to that of CLASSIC: It emphasizes the strong expressivity of the language, at the expense of completeness. The BACK system ([23, 24]) is a formal, well defined system, which is more expressive than CLASSIC, with an incomplete subsumption algorithm. 2] reports on experiments in approximating completeness in an expressive DL. Most presentations in [28] emphasize the need for extending the expressive power and inferential ....

K.V. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The anatomy of the back system. Technical Report KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1987.

....since users of a restricted DL knowledge base use ad hoc methods that jeopardize the soundness of the whole system ( 11] The LOOM system ( 26] takes a contrary approach to that of CLASSIC: It emphasizes the strong expressivity of the language, at the expense of completeness. The BACK system ([23, 24]) is a formal, well defined system, which is more expressive than CLASSIC, with an incomplete subsumption algorithm. 2] reports on experiments in approximating completeness in an expressive DL. Most presentations in [28] emphasize the need for extending the expressive power and inferential ....

K.V. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The back system. Technical Report KIT Report 29, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1985.

....are widely used in many application areas, to date they have focused on representing structured conceptual descriptions, or concepts. A critique of contemporary TKR which argues for greater expressiveness is [Doyle and Patil, 1991] One limitation of current terminological systems, e.g. BACK [von Luck et al. 1987], CLASSIC [Borgida et al. 1989] K Rep [Mays et al. 1991b] KRIS [Baader and Hollunder, 1991] and LOOM [MacGregor and Bates, 1987] is their inability to represent and reason with complex compositions of concepts such as constraint networks where each node is described by a concept. Plans are ....

....with structured plan descriptions. 3.1.1 Concept Languages and their Semantics Terminological knowledge representation, which originated with KL ONE [Brachman and Schmolze, 1985] is an object centered approach in the tradition of semantic networks and frames. Contemporary systems include BACK [von Luck et al. 1987], CLASSIC [Borgida et al. 1989] K Rep [Mays et al. 1991b] KRIS [Baader and Hollunder, 1991] and LOOM [MacGregor and Bates, 1987] Terminological systems share several distinguishing characteristics relevant to our discussion: 1. They are intended to support the definition of conceptual ....

K. von Luck, B. Nebel, C. Pelatson, and A. Schmiedel. The anatomy of the back system. Technical Report KIT- Report 41, Technische Universitat Berlin, Berlin, 1987.

....we refer to them as deductive pattern matchers. In the classification based approach to pattern matching, an instance is matched to a pattern by first abstracting it, and then classifying the abstraction. This strategy is employed by two recent systems, KL TWO [Vilain, 1985] and BACK [Luck et al. 1987; Nebel and Luck, 1987] In this paper, we describe an extension of this approach which (1) is deductively more powerful than, and (2) is expected to be more efficient than, the strategies used in these earlier systems. In many KL ONE style knowledge representation systems, two languages are ....

....as deductive pattern matchers. In the classification based approach to pattern matching, an instance is matched to a pattern by first abstracting it, and then classifying the abstraction. This strategy is employed by two recent systems, KL TWO [Vilain, 1985] and BACK [Luck et al. 1987; Nebel and Luck, 1987]. In this paper, we describe an extension of this approach which (1) is deductively more powerful than, and (2) is expected to be more efficient than, the strategies used in these earlier systems. In many KL ONE style knowledge representation systems, two languages are provided for expressing ....

K. von Luck, B. Nebel, C. Peltason, and A. Schmiedel. The Anatomy of the BACK System. Technical Report KIT Report 41, Technische Universitat Berlin, January 1987.

....can no longer be classified automatically. Description logics evolved from the highly influential kl one system [Brachman and Schmolze, 1985] and its immediate successor, nikl [Schmolze and Mark, 1991] Description logics enjoy a wide variety of contemporary implementations, including back [von Luck et al. 1987], classic [Borgida et al. 1989] crack [Bresciani et al. 1995] k rep [Mays et al. 1991a] kris [Baader and Hollunder, 1991] loom [MacGregor, 1991b] and sb one [Kobsa, 1991] It has seen a correspondingly diverse range of applications, e.g. financial marketing expertise [Apt e et al. ....

K. von Luck, B. Nebel, C. Peltason, and A. Schmiedel. The anatomy of the back system. Technical Report KIT- Report 41, Technische Universitat Berlin, Berlin, 1987.

....since users of a restricted DL knowledge base use ad hoc methods that jeopardize the soundness of the whole system ( 20] The LOOM system ( 37] takes a contrary approach to that of CLASSIC: It emphasizes the strong expressivity of the language, at the expense of completeness. The BACK system ([34,35]) is a formal, well defined system, which is more expressive than CLASSIC, with an incomplete subsumption algorithm. 8] reports on experiments in approximating completeness in an expressive DL. There is a growing agreement that DLs should be strengthened to meet more faithfully users ....

K. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The anatomy of the back system. Technical Report KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1987.

....since users of a restricted DL knowledge base use ad hoc methods that jeopardize the soundness of the whole system ( 20] The LOOM system ( 37] takes a contrary approach to that of CLASSIC: It emphasizes the strong expressivity of the language, at the expense of completeness. The BACK system ([34,35]) is a formal, well defined system, which is more expressive than CLASSIC, with an incomplete subsumption algorithm. 8] reports on experiments in approximating completeness in an expressive DL. There is a growing agreement that DLs should be strengthened to meet more faithfully users ....

K. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The back system. Technical Report KIT Report 29, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1985.

....It also implies pr 2 mechanical product, and dp 2 risky place. The first results from pl being an instance of mechanical plant, and producing pr. The latter holds since w is a toxic waste ( a2) t7) that is buried at dp (a4) In the normalize compare approach, used in CLASSIC ( 7] and BACK ([28, 29]) the knowledge base is kept as a classified taxonomy of normalized concept descriptions. A query is answered by augmenting the taxonomy with the concepts of the query, and checking for their relative positions in the taxonomy. A different, constraints based approach ( 39] KRIS [2] CRACK [13] ....

....the rules are simply F Logic rules, and are interpreted by semantic structures of F Logic. 2.2.1 Rules in Existing Systems In this subsection we shortly describe specific rule sets that were studied in the literature, or implemented in description based systems. 1. BACK and CLASSIC: The BACK ([28, 29]) and CLASSIC ( 7] systems allow rules of the form: X 2 c 0 Gamma X 2 c 1 In CLASSIC, the antecedent, i.e. c 1 must be an already classified concept description. The rules in both systems do not apply to the terminology, and are not made part of the subsumption reasoning. They are used in a ....

K. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The anatomy of the back system. Technical Report KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1987.

....It also implies pr 2 mechanical product, and dp 2 risky place. The first results from pl being an instance of mechanical plant, and producing pr. The latter holds since w is a toxic waste ( a2) t7) that is buried at dp (a4) In the normalize compare approach, used in CLASSIC ( 7] and BACK ([28, 29]) the knowledge base is kept as a classified taxonomy of normalized concept descriptions. A query is answered by augmenting the taxonomy with the concepts of the query, and checking for their relative positions in the taxonomy. A different, constraints based approach ( 39] KRIS [2] CRACK [13] ....

....the rules are simply F Logic rules, and are interpreted by semantic structures of F Logic. 2.2.1 Rules in Existing Systems In this subsection we shortly describe specific rule sets that were studied in the literature, or implemented in description based systems. 1. BACK and CLASSIC: The BACK ([28, 29]) and CLASSIC ( 7] systems allow rules of the form: X 2 c 0 Gamma X 2 c 1 In CLASSIC, the antecedent, i.e. c 1 must be an already classified concept description. The rules in both systems do not apply to the terminology, and are not made part of the subsumption reasoning. They are used in a ....

K. Luck, B. Nebel, C. Peltason, and A. Schmiesel. The back system. Technical Report KIT Report 29, Department of Computer Science, Technische Universitat Berlin, Berlin, FRG, 1985.

....lies at the confluence of several different approaches to the problem of managing information. To begin with, classic belongs to the family of knowledge representation languages that are descendants of klone [ 9 ] including languages and systems such as kandor [ 23 ] loom [ 20 ] and back [ 29 ] . These systems all offer sublanguages for defining terms such that the subsumption relationship holds by definition. The classic concept language resembles them to a considerable extent, but differs from them in one or more of the following features: enumeration using ONE OF, integration of host ....

von Luck, K., Nebel, B., Peltason, C., and Schmiedel, A. The anatomy of the BACK system. KIT-Report 41, Technische Universitat Berlin, January 1987.

....More generally, together with an earlier paper of Brachman and Levesque it shows that terminological reasoning is intractable for any system using a non trivial description language. Finally, consequences of this distressing result are briefly discussed. 1 Introduction The BACK system 1 [13] belongs to the class of hybrid knowledge representation systems based on KL ONE (cf. the article by Brachman and Schmolze [4] As in any other system of this family, a frame based description language (henceforth FDL) which can be viewed as a linear representation of structural inheritance ....

Luck, K. von, Nebel, B., Peltason, C. and Schmiedel, A., The Anatomy of the BACK System, KIT Report 41, Fachbereich Informatik, Technische Universitat Berlin, Berlin, West-Germany, 1987.

....of STP (G; fTg; 1I) Then, either a T x ff is an inequality of the form Gammax 1 e Gamma1 for some edge e 2 E or a e 0 holds for all e 2 E. It is somewhat surprising that, for the Steiner tree packing polyhedron, even it is fulldimensional, a similiar statement is no longer true. In [M92] the interested reader can find an example of a facet defining inequality with positive and negative coefficients. Lemma 5.2 Let G = V; E) be the complete graph on node set V and let N = fT 1 ; TN g; N 2, be a disjoint net list. Furthermore, let a T x ff be a nontrivial ....

....proofs that the inequalities are facet defining. In particular, to prove that the corresponding inequalities are facet defining requires essentially the same scheme. We illustrate this scheme in one sample. For specific proofs of the remaining statements we refer the interested reader to [M92]. Alternating Cycle Inequalities Definition 6.1 Let G = V; E) be a graph and N = fT 1 ; T 2 g a net list. We call a cycle F an alternating cycle with respect to T 1 ; T 2 , if F [T 1 : T 2 ] and V (F ) T 1 T 2 = see Figure 4) Moreover, let F 1 E(T 2 ) and F 2 E(T 1 ) be two sets ....

[Article contains additional citation context not shown here]

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

....331 341 343 switchbox Table 4: The results are quite different for different instances. For two of the examples the wiring length in the Manhattan model is just the same as in the knock knee model though the solutions reported in [GMW92a] have knock knees indeed (for pictures of the solutions, see [M92]) For three other problem instances the wiring length in the Manhattan model exceeds that in the knock knee model by a small amount (for difficult switchbox by 5 ( 1:1 ) for more difficult switchbox by 9 ( 2:0 ) and for pedagogical switchbox by 12 ( 3:6 ) Of course, the shorter ....

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

....will briefly summarize the main ideas for separating the classes of inequalities presented in section 2. The separation algorithms and the associated correctness proofs are quite complicated. In order not to be beyond the scope of this paper, we despense with the proofs and refer the reader to [M92] and [GMW93] for a detailled discussion of this issue. Formally, the separation problem for a given class of inequalities can be stated as follows. Given an instance (G; N ; c) of the Steiner tree packing problem, a vector y 2 IR N ThetaE ; y 0, and a class of valid inequalities for STP (G; N ....

....according to which the choice is made. Besides others, these criteria depend on the location of the terminals of the other nets, the position of the not yet connected terminals of the same net and, again, on the solution y. For a detailed description of these criteria we refer the reader to [M92]. If it is possible to connect the two components on a shortest path by taking the mentioned criterion into account, we connect these two components and choose the next pair of components. Otherwise, we recompute the function f and the sequence by taking the already connected components into ....

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

....0 1 for some i; j 2 f1; lg with bd(t 2 i ; t 2 j ) 1, then t 2 i and t 2 j are not connected in (V; M [ B [ F 2 ) n S 0 1 ) Intuitively, Property (A) becomes clear by drawing a picture. A formal proof of this statement is quite technical and we omit it here (for details see [M92]) Property (B) directly follows from (A) and from Property (iii) Thus, in order to prove that a T x jBj is valid, we can assume that P already satisfies Properties (A) and (B) Now let s 1 denote the number of (connected) components of (T 2 ; B S 1 ) and let s 2 denote the number of ....

....2; 3. This proves that b k e = fi, and the result follows. In fact, finding the appropriate Steiner tree packings as necessary is (somehow) straight forward, but the description of the constructions is quite technical, so we omit the details here. A complete proof of Theorem 4. 4 can be found in [M92]. One of the requirements in Theorem 4.4 is that the net list N is disjoint. One can drop this assumption and still get facet defining inequalities. In this case, however, the edge sets F 2 and F 3 must be extended. The following corollary describes one such case. We state this without a proof and ....

[Article contains additional citation context not shown here]

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

....described above can be adapted to this generalization. It also provides a lower bound for the slack of the most violated inequality. A detailed description of this algorithm requires many technicalities that we do not want to present here. For a discussion of this separation algorithm we refer to [M92]. 5 Determining Cheapest Paths with Costfree Edges The following combinatorial optimization problem is an interesting variant of the shortest path problem. We are given a graph G = V; E) with costs c e 0 for all edges c e 2 E, two nodes s; t 2 V , and a nonnegative integer k. We want to find ....

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

....again, since it is equivalent to solving the Steiner tree packing problem for the instance (G 0 ; N ; c 0 ) where G 0 : V; E 0 : fe 2 E j x e 0g) and c 0 e : x e for e 2 E 0 . A study of the relationship between the polytopes (4.1) and (4. 2) was addressed by Martin [M92]. In particular, he showed that in case where the capacities on the edges are neglected (c = 1) a complete description of STP (G; N ; 1) is given by the facets of the single Steiner tree polyhedra STP (G; fT k g; 1) for k = 1; N . This situation does not hold for STP p (G; N ; 1) ....

A. Martin: Packen von Steinerbaumen: Polyedrische Studien und Anwendung, Ph.D. Thesis, Technische Universitat Berlin, 1992.

.... Bobrow, 1980; Sondheimer and Nebel, 1986] and computer configuration [Owsnicki Klewe, 1988] Based on these ideas, a number of system were built, e.g. kandor [PatelSchneider, 1984] kl two [Vilain, 1985; Schmolze, 1989] krypton [Brachman et al. 1985] meson [Edelmann and Owsnicki, 1986] back [von Luck et al. 1987; Nebel and von Luck, 1988] loom [MacGregor, 1988] classic [Brachman et al. 1989; Borgida et al. 1989] and sb one [Kobsa, 1989] and the formal properties of these systems were investigated [Schmolze and Israel, 1983; Brachman and Levesque, 1984; Patel Schneider, 1986; Levesque and Brachman, ....

Kai von Luck, Bernhard Nebel, Christof Peltason, and Albrecht Schmiedel. The anatomy of the BACK system. KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, West Germany, January 1987.

.... that is, sound and correct with respect to its formalization and, most importantly, that it should give an answer in a reasonable amount 8 Note that this example can even blow up terminological systems which are deliberately designed to be incomplete in their reasoning, e.g. nikl [10] back [30], loom [17] and sb one [13] The only requirement is that the equivalence 8R: C u 8R: D T 8R: C u D) is handled completely. Terminological Reasoning 12 of time. The latter is particularly important when the representation system provides services to a larger system, which depends on timely ....

K. von Luck, B. Nebel, C. Peltason, and A. Schmiedel, The anatomy of the BACK system, KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, West Germany (1987).

....Representation and Reasoning Systems There are a number of examples of hybrid KRR systems described in the literature. First, there is a group of logic based hybrid systems, for instance, the cake system [52] the family of TBox ABox systems (e.g. kl two [60] krypton [9] kandor [48] back [61], meson [14] loom [29] and classic [6] theorem provers using sorted deduction [10, 17] or theory resolution [58] and constraint logic programming [22] Second, we have a group of hybrid systems which integrate logic with other formalisms [16, 51] In the following, we will concentrate on ....

K. von Luck, B. Nebel, C. Peltason, and A. Schmiedel. The Anatomy of the BACK System. KIT Report 41, Department of Computer Science, Technische Universitat Berlin, Berlin, West Germany, Jan. 1987.

.... Schmolze 85] derivates as formalisms for representing terminological knowledge (TBoxes) with formalisms for representing assertions about the actual state of the world (ABoxes) has been investigated 3 (e.g. KRYPTON [Brachman et al. 85] KL TWO [Vilain 85] MESON [Edelmann, Owsnicki 86] BACK [Luck et al. 87] and KANDOR [Patel Schneider 84] The main points in this research were the design of an appropriate ABox, sometimes requiring a restriction of the TBox (e.g. in KRYPTON) and developing means for connecting the reasoning of TBox and ABox. In the sequel we will present one particular solution ....

....The JUNK team cannot exist in this configuration and simultaneously be called a modern small team, or otherwise, KIM cannot a member of the male team. While the arguments above are only informal, it is, of course, possible to formalize the problem using the common semantics of the formalism [Luck et al. 87] and deduce a contradiction as follows: 8x: man(x) human(x) 1) 8x: woman(x) human(x) 2) 8x: man(x) woman(x) 3) 8x: team(x) set(x) 8y: member(x; y) human(y) 8z: leader(x; z) member(x; z) 4) 8x: male team(x) team(x) 8y: member(x; y) man(y) 5) 8x: small team(x) ....

Kai von Luck, Bernhard Nebel, Christof Peltason, Albrecht Schmiedel, The Anatomy of the BACK System, KIT Report 41, Technische Universitat Berlin, Berlin (Germany), January 1987.

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