Organizations need to build and use knowledge to remain viable in the face of competition and change. Due to the limits of organization and the bounded rationality of individuals working within the organizations it is impossible to make all of the required knowledge accessible in explicit and readily retrievable formats. Much of the knowledge an organization needs is held tacitly by members of ...

Reinforcement learning agents typically require a significant amount of data before performing well on complex tasks. Transfer learning methods have made progress reducing sample complexity, but they have primarily been applied to model-free learning methods, not more data-efficient model-based learning methods. This paper introduces timbrel, a novel method capable of transferring information e...

The transfer of medical records is a task all physicians must deal with constantly in daily practice. How often do physicians request medical information from their new patient’s former physician and receive a pile of illegible, poorly organized progress notes along with copies of hospital, emergency, or laboratory records that they do not need? The answer to this question is often. Further, be...

We propose a cross-lingual framework for learning coreference resolvers for resource-poor target languages, given a resolver in a source language. Our method uses word-aligned bitext to project information from the source to the target. To handle task-specific costs, we propose a softmax-margin variant of posterior regularization, and we use it to achieve robustness to projection errors. We sho...