Many inductive logicprogramming systems have operators reorganizing the program so far inferred, such as the intra-construction operator of CIGOL. At the same time, there is a similar reorganizing operator, called th...
详细信息
Many inductive logicprogramming systems have operators reorganizing the program so far inferred, such as the intra-construction operator of CIGOL. At the same time, there is a similar reorganizing operator, called the "folding rule," developed in program transformation. We argue that there are advantages in using an extended folding rule as a reorganizing operator for inductive-inference systems. Such an extended folding rule allows an inductive-inference system not only to recognize already-learned concepts, but also to increase the efficiently of execution of inferred programs.
Recently, strong equivalence for Answer Set programming has been studied intensively, and was shown to be beneficial for modular programming and automated optimization. In this paper we define the novel notion of stro...
详细信息
Recently, strong equivalence for Answer Set programming has been studied intensively, and was shown to be beneficial for modular programming and automated optimization. In this paper we define the novel notion of strong equivalence for logic programs with preferences. Based on this definition we give, for several semantics for preference handling, necessary and sufficient conditions for programs to be strongly equivalent. these results provide a clear picture of the relationship of these semantics with respect to strong equivalence, which differs considerably from their relationship with respect to answer sets. Finally, based on these results, we present for the first time simplification methods for logic programs with preferences.
作者：
Morozov, AARAS
Inst Radio Engn & Elect Moscow 125009 Russia
One of the most interesting and promising approaches to programming Internet agents is logicprogramming of agents. this approach has good prospects, because the ideology and principles of logicprogramming are very c...
ISBN:
(纸本)3540206426
One of the most interesting and promising approaches to programming Internet agents is logicprogramming of agents. this approach has good prospects, because the ideology and principles of logicprogramming are very convenient for searching, recognition, and analysis of unstructured, poorly structured, and hypertext information. Many ideas and methods of logicprogramming of Internet agents based on various modifications of Prolog and non-classical logic (linear, modal, etc.) were developed during the recent decade. Nevertheless, there has been no mathematical apparatus providing sound and complete operation of logic programs in the dynamic Internet environment (i.e., under conditions of permanent update and revision of information). To solve this problem, we have created a mathematical apparatus based on the principle of repeated proving of sub-goals (so-called logical actors).
this papers deals with modelling of the zebra puzzle (also known as Einstein's riddle) in Prolog, the logical programming language. It shows how to construct a rudimentary representation of the puzzle in Prolog an...
详细信息
ISBN:
(纸本)9781467379397
this papers deals with modelling of the zebra puzzle (also known as Einstein's riddle) in Prolog, the logical programming language. It shows how to construct a rudimentary representation of the puzzle in Prolog and then proceeds by suggesting several modifications that significantly reduce the amount of computation time required to solve it. the results illustrate why - if a correct and reasonably efficient knowledge representation is to be constructed - it may be vital for the knowledge engineer to have a firm understanding of the way in which their knowledge representation tool reasons. Naturally, the paper also includes the answer to the eternal question: who owns the zebra?
Justification is the process of computing an evidence for the truth or falsity of an answer to a query in a logic program. there are two well known approaches for computing the evidence: Post-processing based techniqu...
ISBN:
(纸本)3540206426
Justification is the process of computing an evidence for the truth or falsity of an answer to a query in a logic program. there are two well known approaches for computing the evidence: Post-processing based techniques that use tabling engine [3,1] and trace-based techniques such as 4-port debuggers.
We consider the problem of identifying equivalence of two knowledge bases which are capable of abductive reasoning. Here, a knowledge base is written in either first-order logic or nonmonotonic logicprogramming. In t...
详细信息
We consider the problem of identifying equivalence of two knowledge bases which are capable of abductive reasoning. Here, a knowledge base is written in either first-order logic or nonmonotonic logicprogramming. In this work, we will give two definitions of abductive equivalence. the first one, explainable equivalence, requires that two abductive programs have the same explainability for any observation. Another one, explanatory equivalence, guarantees that any observation has exactly the same explanations in each abductive framework. Explanatory equivalence is a stronger notion than explainable equivalence. In first-order abduction, explainable equivalence can be verified by the notion of extensional equivalence in default theories. In nonmonotonic logic programs, explanatory equivalence can be checked by means of the notion of relative strong equivalence. We also show the complexity results for abductive equivalence.
Program changes take up a substantial part of the entire programming effort. Often a preliminary step of improving the design without altering the external behaviour can be recommended. this is the idea behind refacto...
ISBN:
(纸本)3540206426
Program changes take up a substantial part of the entire programming effort. Often a preliminary step of improving the design without altering the external behaviour can be recommended. this is the idea behind refactoring, a source-to-source program transformation that recently came to prominence in the OO-community [1]. Unlike the existing results on automated program transformation, refactoring does not aim at transforming the program entirely automatically. the decision on whether the transformation should be applied and how it should be done is left to the program developer. However, providing automated support for refactoring is useful and an important challenge.
We equip ordered logic programs with negation as failure, using a simple generalization of the preferred answer set semantics for ordered programs. this extension supports a convenient formulation of certain problems,...
详细信息
ISBN:
(纸本)3540206426
We equip ordered logic programs with negation as failure, using a simple generalization of the preferred answer set semantics for ordered programs. this extension supports a convenient formulation of certain problems, which is illustrated by means of an intuitive simulation of logicprogramming with ordered disjunction. the simulation also supports a broader application of "ordered disjunction", handling problems that would be cumbersome to express using ordered disjunction logic programs. Interestingly, allowing negation as failure in ordered logic programs does not yield any extra computational power: the combination of negation as failure and order can be simulated using order (and true negation) alone.
We report on a successful experiment of computer-aided theorem discovery in the area of logicprogramming with answer set semantics. Specifically, withthe help of computers, we discovered exact conditions that captur...
详细信息
We report on a successful experiment of computer-aided theorem discovery in the area of logicprogramming with answer set semantics. Specifically, withthe help of computers, we discovered exact conditions that capture the strong equivalence between a set of a rule and the empty set, a set of a rule and another set of a rule, a set S of two rules and a subset of S with one rule, a set of two rules and a set of another rule, and a set S of three rules and a subset of S with two rules. We prove some general theorems that can help us verify the correctness of these conditions, and discuss the usefulness of our results in program simplification.
We present a higher-order term indexing strategy based on substitution trees. the strategy is based in linear higher-order patterns where computationally expensive parts are delayed. Insertion of terms into the index ...
详细信息
ISBN:
(纸本)3540206426
We present a higher-order term indexing strategy based on substitution trees. the strategy is based in linear higher-order patterns where computationally expensive parts are delayed. Insertion of terms into the index is based on computing the most specific linear generalization of two linear higher-order patterns. Retrieving terms is based on matching two linear higher-order patterns. this indexing structure is implemented as part of the Twelf system to speed-up the execution of the tabled higher-logicprogramming interpreter. Experimental results show substantial performance improvements, between 100% and over 800%.
暂无评论