Learning Rules with Adaptor Grammars

Google Tech Talk
July 6, 2009

ABSTRACT
[note: apologies for the overscanned slides - you can find full resolution slides at http://www.cog.brown.edu/~mj/papers/johnson09-learning-rules-g.pdf ]

Presented by Mark Johnson, Brown University.

Nonparametric Bayesian methods are interesting because they may provide a way of learning the appropriate units of generalization as well as the generalization's probability or weight. Adaptor Grammars are a framework for stating a variety of hierarchical nonparametric Bayesian models, where the units of generalization can be viewed as
kinds of PCFG rules. This talk describes the mathematical and computational properties of Adaptor Grammars and linguistic applications such as word segmentation and syllabification,
and describes the MCMC algorithms we use to sample them.

Joint work with Sharon Goldwater and Tom Griffiths.

Category:

Science & Technology

Tags:

• google
• tech
• talk
• adaptor
• grammars
• computational
• linguistics

[ZirconCode]

Why didn't anyone introduce him? And why was the screen not scaled correctly? The wrong screen scaling ruined the entire thing which was actually quite interesting. Lazy ass editors =(

[bwkaplan]

I agree. They should've used the p-i-p technique like they usually do when showing both speaker and slides at the same time.

[ZirconCode]

At least they provided the slides =/

[Kunado]

First!

[Kunado]

First?