Analogical reasoning is useful to exploit knowledge about similar predicates to define new ones. This paper presents MARs1, a tool that supports the definition of new Prolog predicates with respect to known ones. Star...
详细信息
Autonomous intelligent agents are playing increasingly important roles in our lives. They contain information about us and start to perform tasks on our behalves. Chatbots are an example of such agents that need to en...
详细信息
Autonomous intelligent agents are playing increasingly important roles in our lives. They contain information about us and start to perform tasks on our behalves. Chatbots are an example of such agents that need to engage in a complex conversations with humans. Thus, we need to ensure that they behave ethically. In this work we propose a hybrid logic-based approach for ethical chatbots.
Learning rules with exceptions may be of interest, especially if the exceptions are not important in some sense. Standard Inductive logic programming (ILP) algorithms and classical first order logic are not well-suite...
详细信息
ISBN:
(纸本)3540273263
Learning rules with exceptions may be of interest, especially if the exceptions are not important in some sense. Standard Inductive logic programming (ILP) algorithms and classical first order logic are not well-suited for managing rules with exceptions. Indeed, a hypothesis that is induced accumulates all the exceptions of the rules contained in it. Moreover, with multiple-class problems, classifying an example in two different classes (even if one is the right one) is not correct, so a rule that contains some exceptions may prevent another rule which has no exception from being useful. This paper proposes a new possibilistic logic framework for weighted ILP. It induces rules which are progressively more and more accurate, and allows us to manage exceptions by controlling their accumulation. In this setting, we first propose an algorithm for learning rules when the background knowledge and the examples are stratified into layers having different levels of priority or certainty. This allows the induction of general but uncertain rules together with more specific and less uncertain rules. A second algorithm is presented, which does not require an initial weighted database, but still learn a default set of rules in the possibilistic setting.
Argumentation has gained popularity in AI in recent years to support several activities and forms of reasoning. This talk will trace back the logic programming and non-monotonic reasoning origins of two well-known arg...
详细信息
ISBN:
(纸本)9783319616605;9783319616599
Argumentation has gained popularity in AI in recent years to support several activities and forms of reasoning. This talk will trace back the logic programming and non-monotonic reasoning origins of two well-known argumentation formalisms in AI (namely abstract argumentation and assumption-based argumentation). Finally, the talk will discuss recent developments in AI making use of computational argumentation, in particular to support collaborative decision making.
logic programming Update Languages were proposed as, an extension of logic programming that allows modeling the dynamics of knowledge bases where both extensional (facts) and intentional knowledge (rules) may change o...
详细信息
ISBN:
(纸本)9783642046858
logic programming Update Languages were proposed as, an extension of logic programming that allows modeling the dynamics of knowledge bases where both extensional (facts) and intentional knowledge (rules) may change over time due to updates. Despite their generality, these languages do not;provide a means to directly access past states of the evolving knowledge. They are limited to so-called Markovian change, i.e. changes entirely determined by the current state. We remedy this limitation by extending the logic programming Update Language EVOLP with LTL-like temporal operators that allow referring to the history of the evolving knowledge base, and show how this can be implemented in a logic programming framework.
We study agents situated in partially observable environments, who do not have the resources to create conformant plans. Instead, they create conditional plans which are partial, and learn from experience to choose th...
详细信息
ISBN:
(纸本)9783540766308
We study agents situated in partially observable environments, who do not have the resources to create conformant plans. Instead, they create conditional plans which are partial, and learn from experience to choose the best of them for execution. Our agent employs an incomplete symbolic deduction system based on Active logic and Situation Calculus for reasoning about actions and their consequences. An Inductive logic programming algorithm generalises observations and deduced knowledge in order to choose the best plan for execution. We show results of using PROGOL learning algorithm to distinguish "bad" plans, and we present three modifications which make the algorithm fit this class of problems better. Specifically, we limit the search space by fixing semantics of conditional branches within plans, we guide the search by specifying relative relevance of portions of knowledge base, and we integrate learning algorithm into the agent architecture by allowing it to directly access the agent's knowledge encoded in Active logic. We report on experiments which show that those extensions lead to significantly better learning results.
.Abstract argumentation and logic programming are two formalisms of non-monotonic reasoning that share many similarities. Previous studies contemplating connections between the two formalisms provided back and forth t...
详细信息
ISBN:
(纸本)9783030867720;9783030867713
.Abstract argumentation and logic programming are two formalisms of non-monotonic reasoning that share many similarities. Previous studies contemplating connections between the two formalisms provided back and forth translations from one to the other and found they correspond in multiple different semantics, but not all. In this work, we propose a new set of five argument labels to revisit the semantic correspondences between abstract argumentation and logic programming. By doing so, we shed light on why the two formalisms are not absolutely equivalent. Our investigation lead to the specification of the novel least-stable semantics for abstract argumentation which corresponds to the L-stable semantics of logic programming.
The hybrid probabilistic programs framework [1] allows the user to explicitly encode both logical and statistical knowledge available about the dependency among the events in the program. In this paper, we extend the ...
详细信息
ISBN:
(纸本)3540243623
The hybrid probabilistic programs framework [1] allows the user to explicitly encode both logical and statistical knowledge available about the dependency among the events in the program. In this paper, we extend the language of hybrid probabilistic programs by allowing disjunctive composition functions to be associated with heads of clauses, and we modify its semantics to make it more suitable to encode real-world applications. The new semantics is a natural extension of standard logic programming semantics. The new semantics of hybrid probabilistic programs also subsumes the implication-based probabilistic approach proposed by Lakshmanan and Sadri [12]. We provide also a sound and complete algorithm to compute the least fixpoint of hybrid probabilistic programs with annotated atomic formulas as rule heads.
One challenge faced by many Inductive logic programming (ILP) systems is poor scalability to problems with large search spaces and many examples. Randomized search methods such as stochastic clause selection (SCS) and...
详细信息
ISBN:
(纸本)3540229418
One challenge faced by many Inductive logic programming (ILP) systems is poor scalability to problems with large search spaces and many examples. Randomized search methods such as stochastic clause selection (SCS) and rapid random restarts (RRR) have proven somewhat successful at addressing this weakness. However, on datasets where hypothesis evaluation is computationally expensive, even these algorithms may take unreasonably long to discover a good solution. We attempt to improve the performance of these algorithms on datasets by learning an approximation to ILP hypothesis evaluation. We generate a small set of hypotheses, uniformly sampled from the space of candidate hypotheses, and evaluate this set on actual data. These hypotheses and their corresponding evaluation scores serve as training data for learning an approximate hypothesis evaluator. We outline three techniques that make use of the trained evaluation-function approximator in order to reduce the computation required during an ILP hypothesis search. We test our approximate clause evaluation algorithm using the popular ILP system Aleph. Empirical results are provided on several benchmark datasets. We show that the clause evaluation function can be accurately approximated.
Event-Condition-Action (ECA) languages are an intuitive and powerful paradigm for programming reactive systems. Usually, important features for an ECA language are reactive and reasoning capabilities, the possibility ...
详细信息
ISBN:
(纸本)354039625X
Event-Condition-Action (ECA) languages are an intuitive and powerful paradigm for programming reactive systems. Usually, important features for an ECA language are reactive and reasoning capabilities, the possibility to express complex actions and events, and a declarative semantics. In this paper, we introduce ERA, an ECA language based on, and extending the framework of logic programs updates that, together with these features, also exhibits capabilities to integrate external updates and perform self updates to its knowledge (data and classical rules) and behaviour (reactive rules).
暂无评论