Ambient Intelligence environments consist of various devices that collect, process, change and share the available context information. The imperfect nature of context, the open and dynamic nature of ambient environments, and the special characteristics of the involved devices have introduced new research challenges in the field of KR. Previous work presented a solution based on an extension of multi-context systems through the use of defeasible reasoning to reason efficiently with conflicts. This paper reports on initial experiences gained from the deployment of contextual defeasible reasoning in real environments. We report on the architecture of an implementation on small devices, present the definition and implementation of two concrete application scenarios, and discuss the performance and issues of scalability of the approach.

Measuring the inconsistency degree of a knowledge base can help us to deal with inconsistencies. Several inconsistency measures have been given under different multi-valued semantics, including 4-valued semantics, 3-valued semantics, LPm and Quasi Classical semantics. In this paper, we first carefully analyze the relationship between these inconsistency measures by showing that the inconsistency degrees under 4-valued semantics, 3-value semantics, LPm are the same, but different from the one based on Quasi Classical semantics. We then consider the computation of these inconsistency measures and show that computing inconsistency measurement under multi-valued semantics is usually intractable. To tackle this problem, we propose two novel algorithms that respectively encode the problems of computing inconsistency degrees under 4-valued semantics (3-valued semantics, LPm) and under Quasi Classical semantics into the partial Max-SAT problems. We implement these algorithms and do experiments on some benchmark data sets. The preliminary but encouraging experimental results show that our approach is efficient to handle large knowledge bases.

We consider a generalized form of planning, possibly involving loops, that arises in nondeterministic domains when ex- plicit strong fairness constraints are asserted over the planning domain. Such constraints allow us to specify the necessity of occurrence of selected effects of nondeterministic actions over domain’s runs. Also they are particularly meaningful from the technical point of view because they exhibit the expressiveness advantage of LTL over CTL in verification. We show that planning for reachability and maintenance goals is EXPTIME-complete in this setting, that is, it has the same complexity as conditional planning in nondeterministic domains (without strong fairness constraints). We also show that within the EXPTIME bound one can solve the more general problems of realizing agent planning programs as well as composition-based planning in the presence of strong fairness constraints.

Galles and Pearl [1998] claimed that ``for recursive models, the causal model framework does not add any restrictions to counterfactuals, beyond those imposed by Lewis's [possible-worlds] framework.''  This claim is shown to be false.  Indeed, the opposite claim is true: recursive models are shown to correspond precisely to a subclass of (possible-world) counterfactual structures.  On the other hand, a slight generalization of recursive models, models where all equations have unique solutions, is shown to be incomparable in expressive power to counterfactual structures, despite the fact that the Galles and Pearl arguments should apply to them as well.  The problem with the Galles and Pearl argument is identified: an axiom that they viewed as irrelevant, because it involved disjunction (which was not in their language), is not irrelevant at all. 

By logical proportion, we mean a statement that expresses a semantical equivalence between two pairs of propositions. In these pairs, each element is compared to the other in terms of similarities and/or dissimilarities. An example of such a proportion is the well known analogical proportion: a is to b as c is to d . Analogical proportions have been recently characterized in logical terms, but there are many other proportions that are worth of interest. Some of them can be related to the analogical pattern, others are related to semantical equivalence between conditional objects and express statements such as a ressembles to b and differs from b in the same way as c with respect to d. We show that there are 5 direct proportions, including the analogical one and 4 others having a conditional object flavor, where the change (if any) from a to b goes in the same direction as the change from c to d (if any), together with 5 reverse proportions obtained by switching c and d. Moreover, there exists only one auto-reverse proportion called paralogy and stating that what a and b have in common, c and d have it as well. It is then established that there is none other proportion than these ones (with the exception of 4 degenerated ones) that satisfies a natural “full identity” requirement. The paper proposes a structured and unified view of these logical proportions and discusses their characteristic properties. It extends previous works where only proportions related to analogy were considered. It also explores the use of these logical proportions in transduction-like inference, where new items are classified on the basis of already classified items without trying to induce a generic model, considering similarities and differences between items only. Taking advantage of different proportions, a transduction procedure is proposed.

In discrete dynamical systems change results from actions. As such, given a set of observations, diagnoses often take the form of posited events that result in the observed behaviour. In this paper we revisit formal characterizations of diagnosis, and their relationship to planning. We do so from both a theoretical and a computational perspective. In particular, we extend the characterization of diagnosis to deal with the case of incomplete information, and rich preferences. We also explore the use of state-of-the-art planning technology for the automated generation of diagnoses. Examining several classes of diagnosis problems, we provide both proof of concept and benchmark experiments, the latter showing superior performance to a leading diagnosis engine. Our findings help support the hypothesis that planning technology holds great promise for efficient generation of diagnoses.

Sequential diagnosis takes measurements of an abnormal system to identify faulty components, where the goal is to reduce the diagnostic cost , defined here as the number of measurements. To propose measurement points, previous work employs a heuristic based on reducing the entropy over a set of diagnoses , which can be impractical when the set of diagnoses is too large. Focusing on a smaller set of probable diagnoses scales the approach but generally leads to increased diagnostic cost. We propose a new diagnostic framework employing three new techniques — a more efficient heuristic for measurement point selection, abstraction-based sequential diagnosis, and component cloning — which scales to large systems with good performance in terms of diagnostic cost.

Existing action calculi provide rich, declarative formalisms for reasoning about actions. BDI-based programming languages like AgentSpeak, on the other hand, are procedural and geared towards practical applications of cognitive agents. In this paper, we close the gap between these two lines of research by integrating action calculi and AgentSpeak programs. Specifically, we develop a new and purely declarative semantics for AgentSpeak, which paves the way for combining this language with any suitable action calculus in a strictly modular fashion. As the main technical result, we prove that the new declarative semantics is correct wrt. the standard operational semantics for AgentSpeak. This provides the basis for a modular integration of a BDI-based agent programming language with sophisticated methods for reasoning about actions.

Anytime search algorithms solve optimisation problems by quickly finding a (usually suboptimal) first solution and then finding improved solutions when given additional time. To deliver an initial solution quickly, they are typically greedy with respect to the heuristic cost-to-go estimate h. In this paper, we show that this low-h bias can cause poor performance if the greedy search makes early mistakes. Building on this observation, we present a new anytime approach that restarts the search from the initial state every time a new solution is found. We demonstrate the utility of our method via experiments in PDDL planning as well as other domains, and show that it is particularly useful for problems where the heuristic has systematic errors.

Forward State Planning with Reachability Heuristics is arguably the most successful approach to Automated Planning up to date. In addition to an estimation of the distance to the goal, relaxed plans obtained with such heuristics provide the search with useful information such as helpful actions and look-ahead states. However, this information is extracted only from the beginning of the relaxed plan. In this paper, we propose using information extracted from the last actions in the relaxed plan to generate intermediate goals backwards. This allows us to use information from previous computations of the heuristic and reduce the depth of the search tree.