Material Detail

Approximate Inference in Natural Language Processing

Approximate Inference in Natural Language Processing

This video was recorded at NIPS Workshops, Whistler 2009. I'll start out by presenting an idealized version of the natural language processing problem of parsing. I will brazenly suggest that most of NLP is reducible to variations on parsing problems. I'll show how dynamic programming solves the idealized version of the problem, both for calculating modes and marginals over parse trees, exploiting some key independence assumptions about the structure of natural language sentences. I will then discuss two approximate inference methods that let us build more powerful models of parsing. Neither comes with strong theoretical guarantees, but both are demonstrated to perform strongly in experiments on real NLP data. The first method builds on the dynamic programming representation, combining max-product and sum-product methods to produce, approximately, the k-best parses and a residual sum over the rest of the parses, useful when incorporating features that violate the usual independence assumptions. Experiments validate the approach with a discriminative model for machine translation. The second method turns a parsing problem instance into a concise integer linear program. Approximate inference is then accomplished using well-known linear program relaxation. This is embedded in a new online learning algorithm that tries to penalize uninterpretable fractional solutions (and therefore inference cost at evaluation time). We show that this approach leads to state-of-the-art parsing performance on seven languages, with improved speed for both exact and approximate inference and no significant performance loss.


  • User Rating
  • Comments
  • Learning Exercises
  • Bookmark Collections
  • Course ePortfolios
  • Accessibility Info

More about this material


Log in to participate in the discussions or sign up if you are not already a MERLOT member.