A number of real-world security scenarios can be cast as a problem of transiting an area guarded by a mobile patroller, where the transiting agent aims to choose its route so as to minimize the probability of encountering the patrolling agent, and vice versa. We model this problem as a two-player zero-sum game on a graph, termed the transit game. In contrast to the existing models of area transit, where one of the players is stationary, we assume both players are mobile. We also explicitly model the limited endurance of the patroller and the notion of a base to which the patroller has to repeatedly return. Noting the prohibitive size of the strategy spaces of both players, we develop single- and double-oracle based algorithms including a novel acceleration scheme, to obtain optimum route selection strategies for both players. We evaluate the developed approach on a range of transit game instances inspired by real-world security problems in the urban and naval security domains.

In this paper, we present an alternative to the Turing Test that has some conceptual and practical advantages. A Winograd schema is a pair of sentences that differ only in one or two words and that contain a referential ambiguity that is resolved in opposite directions in the two sentences. We have compiled a collection of Winograd schemas, designed so that the correct answer is obvious to the human reader, but cannot easily be found using selectional restrictions or statistical techniques over text corpora. A contestant in the Winograd Schema Challenge is presented with a collection of one sentence from each pair, and required to achieve human-level accuracy in choosing the correct disambiguation.

Question ambiguity is one major factor that affects question quality. Less ambiguous questions can be produced by using more specific question words. We attack the problem of how to ask more specific questions by supplementing question words with the hypernyms for answer phrases. This dramatically increases the coverage of generated "which" questions. Evaluation results show improved question quality when the question words are disambiguated correctly given the context.

Attitudes play a significant role in determining how individuals process information and behave. In this paper we have developed a new computational model of population wide attitude change that captures the social level: how individuals interact and communicate information, and the cognitive level: how attitudes and concept interact with each other. The model captures the cognitive aspect by representing each individuals as a parallel constraint satisfaction network. The dynamics of this model are explored through a simple attitude change experiment where we vary the social network and distribution of attitudes in a population.

Colorizer is a program that hypothesizes color values that represent a given word or sentence, taking into account both physical descriptions of objects and their emotional connotations. This new application of common sense reasoning uses background knowledge about the world to build a model of the connections between everyday things, and uses this model to guess an appropriate color for a word. Colorizer can run over either static text or real time input, such as a speech recognition stream. It has applications in games, the arts, and webpage design.

This paper presents an agent based framework for modeling of the formation and dynamics of pirate networks. The framework consists of (1) development of network formation mechanism and (2) formulation of pirate attack dynamics. Accordingly, the paper attempts to define the characteristics of Pirate Networks and to formulate the rules that govern the operation and evolution of Pirate Networks. We discuss the clan based social system that facilitate pirate formation as well as the pirate network inter-action with the hosting clan system. Using published material, empirical data and surveys the paper attempts to establish credible formation mechanism and operational characterization of pirate attacks. The proposed framework accounts for clan dynamics and the interplay of social, ecological and physical spaces. Finally we conclude with a discussion on exploratory modeling for the refinement of the proposed framework and for empirically grounding proposed simulations.

Numerous logic frameworks have been proposed for modeling uncertainty and reasoning with such data. While different in syntax, the approaches of these frameworks have been classified into "annotation based" (AB) and "implication based" (IB). In this paper, we present a unified framework which allows evaluating programs in either approach. It extends existing query processing techniques to handle certainty constraints and uses heuristics to further improve the performance. Our experiments indicate that the proposed techniques yield useful tools for uncertainty reasoning.

In this paper, we propose to use recurrent reinforcement learning to directly optimize such trading system performance functions, and we compare two different reinforcement learning methods. The first, Recurrent Reinforcement Learning, uses immediate rewards to train the trading systems, while the second (Q-Learning (Watkins 1989)) approximates discounted future rewards. These methodologies can be applied to optimizing systems designed to trade a single security or to trade portfolios . In addition, we propose a novel value function for risk-adjusted return that enables learning to be done online: the differential Sharpe ratio. Trading system profits depend upon sequences of interdependent decisions, and are thus path-dependent. Optimal trading decisions when the effects of transactions costs, market impact and taxes are included require knowledge of the current system state. In Moody, Wu, Liao & Saffell (1998), we demonstrate that reinforcement learning provides a more elegant and effective means for training trading systems when transaction costs are included, than do more standard supervised approaches.

In this paper, we propose to use recurrent reinforcement learning to directly optimize such trading system performance functions, and we compare two different reinforcement learning methods. The first, Recurrent Reinforcement Learning, uses immediate rewards to train the trading systems, while the second (Q-Learning (Watkins 1989)) approximates discounted future rewards. These methodologies can be applied to optimizing systems designed to trade a single security or to trade portfolios . In addition, we propose a novel value function for risk-adjusted return that enables learning to be done online: the differential Sharpe ratio. Trading system profits depend upon sequences of interdependent decisions, and are thus path-dependent. Optimal trading decisions when the effects of transactions costs, market impact and taxes are included require knowledge of the current system state. In Moody, Wu, Liao & Saffell (1998), we demonstrate that reinforcement learning provides a more elegant and effective means for training trading systems when transaction costs are included, than do more standard supervised approaches.

This article analyzes an attempt to use computing technology, including AI, to improve the combat readiness of a U.S. Navy aircraft carrier. The method of introducing new technology, as well as the reaction of the organization to the use of the technology, is examined to discern the reasons for the rejection by the carrier's personnel of a technically sophisticated attempt to increase mission capability. This effort to make advanced computing technology, such as expert systems, an integral part of the organizational environment and, thereby, to significantly alter traditional decision-making methods failed for two reasons: (1) the innovation of having users, as opposed to the navy research and development bureaucracy, perform the development function was in conflict with navy operational requirements and routines and (2) the technology itself was either inappropriate or perceived by operational experts to be inappropriate for the tasks of the organization. Finally, this article suggests those obstacles that must be overcome to successfully introduce state-of-the-art computing technology into any organization.