We use graphical models to explore the question of how people learn simple causal relationships from data. The two leading psychological theories can both be seen as estimating th...
We argue that multilingual parallel data provides a valuable source of indirect supervision for induction of shallow semantic representations. Specifically, we consider unsupervi...
Adaptor grammars extend probabilistic context-free grammars to define prior distributions over trees with "rich get richer" dynamics. Inference for adaptor grammars seek...
Agent training techniques study methods to embed empirical, inductive knowledge representations into intelligent agents, in dynamic, recursive or semi-automated ways, expressed in...