The arc-factored model is the simplest graph-based model for dependency parsing, a model that factorizes the scoring function in terms of subgraphs of the dependency tree. The arc-factored model is often referred to as a first-order model, because it decomposes the score of a tree into the scores of one arc at a time. From a linguistic point of view, this is a rather drastic assumption, similar...

Compiling Bayesian networks (BNs) is one of the hot topics in the area of probabilistic modeling and processing. In this paper, we propose a new method of compiling BNs into multi-linear functions (MLFs) based on Zero-suppressed BDDs (ZBDDs), which are the graph-based representation of combinatorial item sets. Our method is different from the original approach of Darwiche et. al which encodes B...

Many solution methods for Markov Decision Processes (MDPs) exploit structure in the problem and are based on value function factorization. Especially multiagent settings, however, are known to suffer from an exponential increase in value component sizes as interactions become denser, restricting problem sizes and types that can be handled. We present an approach to mitigate this limitation for ...

Envisioning has been used extensively to model behavior of physical systems. Envisioning generates the qualitatively distinct possible behaviors without numerically simulating every possible set of input conditions and model parameters. This paper applies envisioning to analyze course of action (COA) diagrams to determine the qualitatively distinct outcomes of military operations. In order to a...