Tim Vieira

Graduate Student

About

I'm a first year PhD student in the CLSP @ Johns Hopkins. Previously, I was a programmer for Andrew McCallum at UMass where I worked on Rexa and factorie. Before that, I was a programmer for Dan Roth at UIUC where my research was primarily focused on Textual Entailment.

Research Interests

keywords: machine learning, natural language processing, information extraction, learning and inference in complex factor graphs, distantly-supervised learning.

I apply machine learning to tough problems—tending toward natural language processing and information extraction tasks.

Fun

Roller Derby

My alter ego, Rambo Sprinkles, plays derby with the Dirty Dozen.

Snakeboarding

I used to be quite the snakeboarder (video).

Cars

The 5th Ave (video) - a gorgeous 1984 Chrysler Fifth Avenue purchased for a mere $900 - installed the hydraulics myself!

The Big Bad Baby Blue Buick (video) - hydraulics on the back this time. You can't tell in this video, but the car was scrapping the ground.

News

• April 15, 2011: I will be joining the PhD program at Johns Hopkins this Fall!

Publications

• M. Sammons, V. Vydiswaran, T. Vieira, N. Johri, M. Chang, D. Goldwasser, V. Srikumar, G. Kundu, Y. Tu, K. Small, J. Rule, Q. Do and D. Roth,
  Relation Alignment for Textual Entailment Recognition. Text Analysis Conference (TAC) (2009)

Code & Resources

Factorie

FACTORIE is a probabilistic programming framework implemented as a software library in Scala. Providing users with a succinct framework for creating abritrary factor graphs with tied parameters, estimating parameters, and performing inference.

Python Utilities

Collection of various utilities, recipes, and "gutted" components of other projects.

Notes

Notes on various subjects, including statistics, sampling, reinforcement learning, and implementation tricks common in machine learning algorithms.