Standard belief contraction assumes an underlying logic containing full classical propositional logic, but there are good reasons for considering contraction in less expressive logics. In this paper we focus on Horn logic. In addition to being of interest in its own right, our choice is motivated by the use of Horn logic in several areas, including ontology reasoning in description logics. We consider three versions of contraction: entailment-based and inconsistency-basedcontraction (e-contraction and i-contraction, resp.), introduced by Delgrande for Horn logic, and package contraction (p-contraction), studied by Fuhrmann and Hansson for the classical case. We show that the standard basic form of contraction, partial meet, is too strong in the Horn case. We define more appropriate notions of basic contraction for all three types above, and provide associated representation results in terms of postulates. Our results stand in contrast to Delgrande's conjectures that orderly maxichoice is the appropriate contraction for both e- and i-contraction. Our interest in p-contraction stems from its relationship with an important reasoning task in ontological reasoning:repairing the subsumption hierarchy in EL. This is closely related to p-contraction with sets of basic Horn clauses (Horn clauses of the form p -> q). We show that this restricted version of p-contraction can also be represented as i-contraction.

Least Common Subsumers (LCS) have been proposed in Description Logics (DL) to capture the commonalities between two or more concepts. Since its introduction in 1992,  LCS have been successfully employed  as a logical tool for a variety of applications, spanning from inductive learning, to bottom-up construction of knowledge bases, information retrieval,  to name a few. The best known algorithm for computing LCS uses structural comparison on normal forms, and the most expressive DL it is applied to is ALEN. We provide a general tableau-based calculus for computing LCS, via substitutions on concept terms containing concept variables. We show the applicability of our method to an expressive DL (but without disjunction and full negation), discuss complexity issues, and show the generality of our proposal.

Probabilistic programming languages allow a modeler to build probabilistic models using complex data structures with all the power of a programming language. We present CTPPL, an expressive probabilistic programming language for dynamic processes that models processes using continuous time. Time is a first class element in our language; the amount of time taken by a subprocess can be specified using the full power of the language. We show through examples that CTPPL can easily represent existing continuous time frameworks and makes it easy to represent new ones. We present semantics for CTPPL in terms of a probability measure over trajectories. We present a particle filtering algorithm for the language that works for a large and useful class of CTPPL programs.

Optimization in graphical models is an important problem which has been studied in many AI frameworks such as weighted CSP, maximum satisfiability or probabilistic networks. By identifying conditionally independent subproblems, which are solved independently and whose optimum is cached, recent Branch and Bound algorithms offer better asymptotic time complexity. But the locality of bounds induced by decomposition often hampers the practical effects of this result because subproblems are often uselessly solved to optimality. Following the Russian Doll Search (RDS) algorithm, a possible approach to overcome this weakness is to (inductively) solve a relaxation of each subproblem to strengthen bounds. The algorithm obtained generalizes both RDS and tree-decomposition based algorithms such as BTD or AND-OR Branch and Bound. We study its efficiency on different problems, closing a very hard frequency assignment instance which has been open for more than 10 years.

Mechanical design tools would be considerably more useful if we could interact with them in the way that human designers communicate design ideas to one another, i.e., using crude sketches and informal speech. Those crude sketches frequently contain pen strokes of two different sorts, one type portraying device structure, the other denoting gestures, such as arrows used to indicate motion. We report here on techniques we developed that use information from both sketch and speech to distinguish gesture strokes from non-gestures -- a critical first step in understanding a sketch of a device. We collected and analyzed unconstrained device descriptions, which revealed six common types of gestures. Guided by this knowledge, we developed a classifier that uses both sketch and speech features to distinguish gesture strokes from non-gestures. Experiments with our techniques indicate that the sketch and speech modalities alone produce equivalent classification accuracy, but combining them produces higher accuracy.

Building and using agent-based models is often impractical, in part due to the cost of including expensive subject matter experts (SMEs) in the development process. In this paper, we describe a method for "bootstrapping" model building to lower the cost of overall model development. The models we are interested in here capture dynamic phenomena related to international and subnational conflict. The method of acquiring these models begins with event data drawn from news reports about a conflict region, and infers model characteristics particular to a conflict modeling framework called the Power Structure Toolkit (PSTK). We describe the toolkit and how it has been used prior to this work. We then describe the current problem of modeling conflict and the empirical data available to learn models, and extensions to the PSTK for model generation from this data. We also describe a formative evaluation of the system that compares the performance and costs of models built entirely by an SME against models built with an SME aided by the automated model generation process. Early results indicate at least equivalent prediction rates with significant savings in model generation costs.

Given an imagebase with tagged images, four types of tasks an be executed, i.e., content-based image retrieval, image annotation, text-based image retrieval, and query expansion. For any of these tasks the similarity on the concerned type of objects is essential. In this paper, we propose a framework to tackle these four tasks from a unified view. The essence of the framework is to estimate similarities by exploiting the interactions between objects of different modality. Experiments show that the proposed method can improve similarity estimation, and based on the improved similarity estimation, some simple methods can achieve better performances than some state-of-the-art techniques.

Sparse coding is an unsupervised learning algorithm for finding concise, slightly higher-level representations of inputs, and has been successfully applied to self-taught learning, where the goal is to use unlabeled data to help on a supervised learning task, even if the unlabeled data cannot be associated with the labels of the supervised task (Raina et al., 2007). However, sparse coding uses a Gaussian noise model and a quadratic loss function, and thus performs poorly if applied to binary valued, integer valued, or other non-Gaussian data, such as text. Drawing on ideas from generalized linear models (GLMs), we present a generalization of sparse coding to learning with data drawn from any exponential family distribution (such as Bernoulli, Poisson, etc). This gives a method that we argue is much better suited to model other data types than Gaussian. We present an algorithm for solving the L1-regularized optimization problem defined by this model, and show that it is especially efficient when the optimal solution is sparse. We also show that the new model results in significantly improved self-taught learning performance when applied to text classification and to a robotic perception task.

Location  plays a very important role in the retail business due to its huge and long-term investment. In this paper, we propose a novel semi-supervised regression model for evaluating convenience store location based on spatial data analysis. First, the input features for each convenience store can be extracted by analyzing the elements around it based on a geographic information system, and the turnover is used to evaluate its performance. Second, considering the practical application scenario, a manifold regularization model with one semi-supervised performance information constraint is provided. The promising experimental results in the real-world dataset demonstrate the effectiveness of the proposed approach  in performance prediction of  certain candidate locations for new convenience store opening.

Almost all approaches to model-based diagnosis presume that the system being diagnosed behaves non-intermittently and analyze behavior over a small number (often only one) of time instants.  In this paper we show how existing approaches to model-based diagnosis can be extended to diagnose intermittent failures as they manifest themselves over time. In addition, we show where to insert probe points to best distinguish among the intermittent faults those that best explain the symptoms and isolate the fault in minimum expected cost.