Eiter, T.; Faber, W.; Leone, N.; Pfeifer, G.; and Polleres, A. 2003. A logic programming approach to knowledge-state planning, II: The DLV K system. Artif. Intell. 144(1--2):157--211.  Home/Search   Document Details and Download   Summary   Related Articles   Check

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A Logic Programming Approach to Knowledge-State Planning: Semantics and Complexity. To appear in ACM TOCL, 2003.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A Logic Programming Approach to Knowledge-State Planning: Semantics and Complexity. ACM TOCL, 2003. To appear.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning: Semantics and complexity. ACM Transactions on Computational Logic, 2003. To appear.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV system. Artificial Intelligence, 144(1-2):157--211, 2002.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV system. Artif. Intell., 144(1-2):157-211, 2003.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledgestate planning: Semantics and complexity. ACM Transactions on Computational Logic, 5(2), April 2004. To appear.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledgestate planning, II: The DLV system. Artificial Intelligence, 144(1--2):157--211, March 2003.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV system. Artif. Intell., 144(12) :157-211, 2003.

No context found.

Eiter, T., Faber, W., Leone, N., Pfeifer, G., Polleres, A.: A Logic Programming Approach to Knowledge-State Planning: Semantics and Complexity, ACM Transactions on Computational Logic, 2003, To appear.

No context found.

Eiter, T., Faber, W., Leone, N., Pfeifer, G., Polleres, A.: A Logic Programming Approach to Knowledge-State Planning, II: The DLV System, Artificial Intelligence, 144(1-2), 2002, 157--211.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV system. Artif. Intell., 144(1-2):157--211, 2003.

....and presents a (here simplified version) of a multi agent system in the postal services domain. After that, in Section 3 we describe how to model the intended behavior of a multi agent system as an abstract planning problem, and instantiate this for our example system using the action language K [5, 4]. Our approach to agent monitoring is then discussed in Section 4, where we also investigate some fundamental properties. After a brief discussion of the implementation in Section 5 and a review of related work in Section 6, we conclude in Section 7 with an outlook on further research. 2 Message ....

....power (loop, conditionals) automated plan generation is limited in this formalism. In Subsection 3.1 we give a generic formulation of our approach, independent of a particular planning mechanism. Then, in Subsection 3. 2 we instantiate this high level description using the action language K [5, 4]. While our approach does not rely on K, we have chosen it because of its declarative nature and its capabilities to deal with incomplete knowledge and nondeterminism. 3.1 Modelling Intended Behavior of a MAS Our approach to formalize the intended collaborative behavior of a MAS consisting of ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning: Semantics and complexity. ACM Transactions on Computational Logic, 2003. To appear.

....and presents a (here simplified version) of a multi agent system in the postal services domain. After that, in Section 3 we describe how to model the intended behavior of a multi agent system as an abstract planning problem, and instantiate this for our example system using the action language K [5, 4]. Our approach to agent monitoring is then discussed in Section 4, where we also investigate some fundamental properties. After a brief discussion of the implementation in Section 5 and a review of related work in Section 6, we conclude in Section 7 with an outlook on further research. 2 Message ....

....power (loop, conditionals) automated plan generation is limited in this formalism. In Subsection 3.1 we give a generic formulation of our approach, independent of a particular planning mechanism. Then, in Subsection 3. 2 we instantiate this high level description using the action language K [5, 4]. While our approach does not rely on K, we have chosen it because of its declarative nature and its capabilities to deal with incomplete knowledge and nondeterminism. 3.1 Modelling Intended Behavior of a MAS Our approach to formalize the intended collaborative behavior of a MAS consisting of ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: the DLV system. Artificial Intelligence, 144(1-2):157--211, 2003.

....language C [8] In addition to allowing for conditional and nondeterministic effects of actions, C also supports concurrent actions as well as indirect effects and preconditions of actions through static causal laws. Closely related to it is the recent planning language K [3]. There are a number of formalisms for probabilistic reasoning about actions. In particular, Bacchus et al. 1] propose a probabilistic generalization of the situation calculus, which is based on first order logics of probability, and which allows to reason about an agent s probabilistic degrees ....

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV K system. Artif. Intell., 144:157--211, 2003.

....be unknown, e.g. whether a door in front of the robot is open) they must take decisions, execute actions, and reason on the basis of their (incomplete) information at hand. For example, if it is not known whether a door is open, the robot might do a sensing action, or decide to push back. In [5,6], we have proposed a new language, K (where K should remind of states of knowledge) for planning under incomplete knowledge. This language is very flexible, and is capable of modeling transitions between states of the world (i.e. states of complete knowledge) and reasoning about them as a ....

....languages, in particular Giunchiglia and Lifschitz action language C [17,26,29] K is closer in spirit to answer set semantics [12] than to classical logics. It supports the explicit use of default negation, and thus exploiting the power of answer sets to deal with incomplete knowledge. In [6] we have defined the syntax and semantics of K, discussed how it can be used for knowledge representation, plus we have analyzed the computational complexity of planning in K. In the present paper, which is Part II of this series of papers, we turn to the DLV planning system, which implements ....

[Article contains additional citation context not shown here]

Eiter, T., Faber, W., Leone, N., Pfeifer, G., Polleres, A., Dec. 2001. A Logic Programming Approach to Knowledge-State Planning: Semantics and Complexity. Tech. Rep. INFSYS RR-1843-01-11, TU Wien.

.... Finally, the goal: section defines the goal to be reached, which is a conjunction of ground fluent literals left of , and the plan length which is given in parentheses right of (in this case, 3) A detailed account of the language K is given in [39] The planning front end, described in [37], is available, ready for experiments, at http: www.dlvsystem.com K . 5.4 SQL3 Front End The front end for SQL3 was the first front end developed for DLV. At the time when it was realized, no system fully supported the SQL3 query language, and this front end allowed the users to play with some ....

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A Logic Programming Approach to Knowledge-State Planning, II: the DLV System. Technical Report INFSYS RR-1843-01-12, Institut fur Informationssysteme, Technische Universitat Wien, Dec. 2001.

....describing that a specific person is on the other side of the river. Here the literals after requires must be classical literals of the static background knowledge (like person(X) and person(Y) or literals of built in predicates (such as X Y) Our implementation of K, the DLV system [8], currently supports the built in predicates A B , A = B , A = B with the obvious meaning of less than, less orequal and inequality for strings and numbers, the arithmetic built ins A = B C and A = B C which stand for integer addition and multiplication, and the predicate #int(X) ....

....or scheduling problems; optimization, however, remains restricted to action costs. 6 Transformation to Logic Programming In this section, we describe how planning under action costs can be implemented by means of a transformation to answer set programming. It extends the transformation given in [8], which maps ordinary K planning problems P to disjunctive logic programs lp(P) under the answer set semantics [16] The extension to this mapping takes advantage of weak constraints, cf. 3, 4] 6.1 Disjunctive logic programs with weak constraints First, we give a brief review of disjunctive ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A Logic Programming Approach to Knowledge-State Planning, II: the DLV System. Technical Report INFSYS RR-1843-01-12, Institut fur Informationssysteme, Technische Universitat Wien, Dec. 2001.

....Figure 1 specifies the arguments of action cross2, where two persons cross the bridge together, while line (4) specifies a fluent describing the fact that a specific person is on the other side of the river. The literals after requires in a declaration are from or built in predicates. DLV [7], our implementation of K, currently supports built in predicates a b , a = b , and a = b with the obvious meaning for strings and numbers, predicates a = b c , a = b c for integer arithmetics, and #int(X) which enumerates all integers (up to a limit set by the user) Causation ....

....two step plan, is: fmove(1; 3) move(2; 4) move(6; 5)g; ffinishg; i 5 Implementation We briefly describe how planning under action costs can be implemented using a translation to answer set programming. We will define an extension lp (P) of the logic program lp(P) as defined in [7], such that its optimal answer sets (i.e. those minimizing weak constraint violation, see [10, 3] correspond to the optimal cost plans for a planning problem P. We recall that in lp(P) fluent and action literals are extended by an additional time parameter, and executability conditions as well ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning, II: The DLV system. Technical Report INFSYS RR-1843-0112, Inst. f. Informationssysteme, TU Wien, December 2001.

....and effectively solved. 1 Introduction Recently, several declarative planning languages and formalisms have been introduced, which allow for an intuitive encoding of complex planning problems including ramifications, incomplete information, non deterministic action effects, or parallel actions [13, 18, 17, 19, 12, 4 6]. While these formalisms are designed to generate any plans that establish the planning goals, in practice we are usually interested in particular plans that are optimal with respect to an objective function which measures the quality (or cost) of a plan. Often, this is just the number of time ....

....cost of the actions in that plan. In answer set planning [18] where plans are represented by answer sets of a logic program, this kind of problem has not been addressed so far, to the best of our knowledge. In this paper, we address this issue and present an extension of the planning language K [5, 6], where one can associate actions with costs. The main contributions are: We define syntax and semantics of a modular extension to K. Costs are associated to an action by a designated where clause describing a cost value. Action costs may be dynamic, as they potentially depend on the ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A logic programming approach to knowledge-state planning: Semantics and complexity. Technical Report INFSYS RR-184301 -11, Inst. f. Informationssysteme, TU Wien, December 2001.

....be unknown, e.g. whether a door in front of the robot is open) they must take decisions, execute actions, and reason on the basis of their (incomplete) information at hand. For example, if it is not known whether a door is open, the robot might do a sensing action, or decide to push back. In [10], we have proposed a new language, K (where K should remind of states of knowledge) for planning under incomplete knowledge. This language is very flexible, and is capable of modeling transitions between states of the world (i.e. states of complete knowledge) and reasoning about them as a ....

....languages, in particular Giunchiglia and Lifschitz action language C [6, 7, 11] K is closer in spirit to answer set semantics [12] than to classical logics. It supports the explicit use of default negation, and thus exploiting the power of answer sets to deal with incomplete knowledge. In [10] we have defined the syntax and semantics of K, discussed how it can be used for knowledge representation, plus we have analyzed the computational complexity of planning in K. In the present paper, which is Part II of this series of papers, we turn to the DLV K planning system, which implements ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, A. Polleres, A Logic Programming Approach to KnowledgeState Planning: Semantics and Complexity, Tech. Rep. INFSYS RR-1843-01-11, Institut fur Informationssysteme, Technische Universitat Wien (Dec. 2001).

....results. In Section 6, we discuss related work, and the final Section 7 discusses further work and draws some conclusions. The present paper is part I in a series of papers which comprehensively describe our work, and contains the foundational semantic definitions and theoretical results; part II [12] reports about the DLV K system (which is freely available at URL:http: www.dbai.tuwien.ac.at proj dlv ) and in particular contains an experimental evaluation and comparisons to other planning systems (for a theoretical account, see also Section 6) 2 Language K In this section, we will ....

....initially : clogged(T) Also in this case the resulting problem domains are deterministic and hence optimistic plans and secure plans coincide. This indicates that planning problems of this section can be solved faster than those of section 3.3. Indeed, we have observed this also experimentally [12]; the encodings of section 3.4 can often be solved several orders of magnitudes faster than those of section 3.3 in the DLV K system prototype. 3.5 Discussion As we have seen in the preceding subsections, the use of knowledge states instead of world states allows us to represent planning ....

[Article contains additional citation context not shown here]

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres. A Logic Programming Approach to Knowledge-State Planning, II: the DLV K System. Manuscript, November 2001.

No context found.

Eiter, T.; Faber, W.; Leone, N.; Pfeifer, G.; and Polleres, A. 2003. A logic programming approach to knowledge-state planning, II: The DLV K system. Artif. Intell. 144(1--2):157--211.

No context found.

T. Eiter, W. Faber, N. Leone, G. Pfeifer, and A. Polleres, `A logic programming approach to knowledge-state planning, II: The DLV K system ', Artif. Intell., 144(1--2), 157--211, (2003).

No context found.

Eiter, T., Faber, W., Leone, N., Pfeifer, G., and Polleres, A., A Logic Programming Approach to Knowledge-State Planning: Semantics and Complexity. ACM Transaction on Computational Logic, 2002, to appear.

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