Mechanisms (especially on the Internet) have begun allowing people or organizations to express richer preferences in order to provide for greater levels of overall satisfaction.  In this paper, we develop an operational methodology for quantifying the expected gains in economic efficiency associated with different forms of expressiveness.  We begin by proving that the sponsored search mechanism (GSP) used by Google, Yahoo!, MSN, etc. can be arbitrarily inefficient. We then experimentally compare its efficiency to a slightly more expressive variant (PGSP), which solicits an extra bid for a premium class of positions. We generate random preference distributions based on published industry knowledge.  We determine ideal strategies for the agents using a custom tree search technique, and we also benchmark using straightforward heuristic bidding strategies. The GSP's efficiency loss is greatest in the practical case where some advertisers ("brand advertisers") prefer top positions while others ("value advertisers") prefer middle positions, and that loss can be dramatic. It is also worst when agents have small profit margins. While the PGSP is only slightly more expressive (and thus not much more cumbersome), it removes almost all of the efficiency loss in all of the settings we study.

Many local search algorithms are based on searching in the k-exchange neighborhood. This is the set of solutions that can be obtained from the current solution by exchanging at most k elements. As a rule of thumb, the larger k is, the better are the chances of  finding an improved  solution. However,  for inputs of size n,  a naive brute-force search of the k-exchange neighborhood requires n (O(k)) time, which is not practical even for very small values of k. We show that for several classes of sparse graphs, like planar graphs, graphs of bounded vertex degree and graphs excluding some fixed graph as a minor, an improved solution in the k-exchange neighborhood for many problems can be found much more efficiently.  Our algorithms run in time O(T(k)*n c ), where T is a function depending on k only and c is a constant independent of k.  We demonstrate the applicability of this  approach on different  problems like r-Center, Vertex Cover,  Odd Cycle Transversal, Max-Cut, and  Min-Bisection.  In particular, on planar graphs, all our algorithms searching for a k-local improvement run in time O(2 (O(k) ) * n (2) ), which is polynomial for k=O(log n). We also complement the algorithms with complexity results indicating that brute force search is unavoidable in more general classes of sparse graphs.

Computing the semantic distance between real-world concepts is crucial for many intelligent applications. We present a novel method that leverages data from `Wikispeedia', an online game played on Wikipedia; players have to reach an article from another, unrelated article, only by clicking links in the articles encountered. In order to automatically infer semantic distances between everyday concepts, our method effectively extracts the common sense displayed by humans during play, and is thus more desirable, from a cognitive point of view, than purely corpus-based methods. We show that our method significantly outperforms Latent Semantic Analysis in a psychometric evaluation of the quality of learned semantic distances.

Most sketch recognition systems are accurate in recognizing either text or shape (graphic) ink strokes, but not both. Distinguishing between shape and text strokes is, therefore, a critical task in recognizing hand drawn digital ink diagrams which commonly contain many text labels and annotations. We have found the ‘entropy rate’ to be an accurate criterion of classification. We found that the entropy rate is significantly higher for text strokes compared to shape strokes and can serve as a distinguishing factor between the two. Using entropy values, our system produced a correct classification rate of 92.06% on test data belonging to diagrammatic domain for which the threshold was trained on.  It also performed favorably on data for which no training examples at all were supplied.

In the context of Dung's theory of abstract argumentation frameworks, the recently introduced resolution-based grounded semantics features the unique property of fully complying with a set of general requirements, only partially satisfied by previous literature proposals. This paper contributes to the investigation of resolution-based grounded semantics by analyzing its computational properties with reference to a standard set of decision problems for abstract argumentation semantics: (a) checking the property of being an extension for a set of arguments; (b) checking agreement with traditional grounded semantics; (c) checking the existence of a non-empty extension; (d) checking credulous acceptance of an argument; (e) checking skeptical acceptance of an argument. It is shown that problems (a)-(c) admit polynomial time decision processes, while (d) is NP-complete and (e) coNP-complete.

We show how to translate HTN domain descriptions (if they satisfy certain restrictions) into PDDL so that they can be used by classical planners.  We provide correctness results for our translation algorithm, and show that it runs in linear time and space. We also show that even small and incomplete amounts of HTN knowledge, when translated into PDDL using our algorithm, can greatly improve a classical planner's performance. In experiments on several thousand randomly generated problems in three different planning domains, such knowledge speeded up the well-known Fast-Forward planner by several orders of magnitude, and enabled it to solve much larger problems than it could otherwise solve.

These existing algorithms discover projections policies in very high dimensional state spaces. of the training data under which nearby points are likely We propose a linear dimensionality reduction algorithm to have the same class label or similar regression targets. The that discovers predictive projections: projections algorithm described in this paper makes use of the same machinery in which accurate predictions of future states but attempts to find low-dimensional projections under can be made using simple nearest neighbor style which current state vectors accurately predict future states learning. The goal of this work is to extend the in the projected space. The intuition is that projections which reach of existing reinforcement learning algorithms capture the state dynamics in this way are likely to contain to domains where they would otherwise be inapplicable information that will be useful for control.

Many problems have a huge state space and no good heuristic to order moves so as to guide the search toward the best positions. Random games can be used to score positions and evaluate their interest. Random games can also be improved using random games to choose a move to try at each step of a game. Nested Monte-Carlo Search addresses the problem of guiding the search toward better states when there is no available heuristic. It uses nested levels of random games in order to guide the search. The algorithm is studied theoretically on simple abstract problems and applied successfully to three different games: Morpion Solitaire, SameGame and 16x16 Sudoku.

In this work we propose gRegress, a new algorithm which given set of labeled graphs and a real value associated with each graph extracts the complete set of subgraphs such that a) each subgraph in this set has correlation with the real value above a user-specified threshold and b) each subgraph in this set has correlation with any other subgraph in the set below a user-specified threshold. gRegress incorporates novel pruning mechanisms based on correlation of a subgraph feature with the output and correlation with other subgraph features. These pruning mechanisms lead to significant speedup. Experimental results indicate that in terms of runtime, gRegress substantially outperforms gSpan, often by an order of magnitude while the regression models produced by both approaches have comparable accuracy.

This paper considers the problem of composing or scheduling several (non-deterministic) behaviors so as to conform to a specified target behavior as well as satisfying constraints imposed by the environment in which the behaviors are to be performed. This problem has already been considered by several works in the literature and applied to areas such as web service composition, the composition of robot behaviors and co-ordination of distributed devices. We develop a sound and complete algorithm for determining such a composition which has a number of significant advantages over previous proposals: a) our algorithm is different from previous proposals which resort to dynamic logic or simulation relations, b) we realized an implementation in Java as opposed to other approaches for which there are no known implementations, c) our algorithm determines all possible schedulers at once, and d) we can use our framework to define a notion of approximation when the target behavior cannot be realized.