David Leo Wright Hall

| papers | thesis | projects | cv
Picture of David

Email: dlwh at cs.[university].edu

I'm a fifth year PhD student in EECS at Berkeley in the Berkeley NLP Group. My advisor is Dan Klein. I'm primarily interested in Natural Language Processing and Machine Learning, particularly computational historical linguistics and statistical parsing. I'm supported by a Google PhD fellowship in Natural Language Processing. Previously, I was supported by an NSF fellowship.

Before that, I was an undergrad in Symbolic Systems at Stanford University, working in the Stanford NLP Group with Dan Jurafsky and Chris Manning. There, I worked on the Mimir Project, which aims to measure the effect of funding and institutional structure on interdisciplinarity and research. I also picked up an M.S. somewhere along the way.

Publications

Less Grammar, More Features[bib][brief] [pdf]
David Hall, Greg Durrett, and Dan Klein
ACL, 2014

Sparser, Better, Faster GPU Parsing[bib][brief] [pdf]
David Hall, Taylor Berg-Kirkpatrick, John Canny, and Dan Klein
ACL, 2014

A multi-Teraflop Constiuency Parser using GPUs[bib][brief][pdf]
John Canny, David Hall, and Dan Klein
EMNLP, 2013

Decentralized Entity-Level Modeling for Coreference Resolution[bib][brief][pdf]
Greg Durrett, David Hall, and Dan Klein
ACL, 2013

Dating Proto-Indo-European: A revised computational analysis supports the steppe hypothesis [bib][brief]
Will Chang, David Hall, Chundra Cathcart, and Andrew Garrett
To appear, International Conference on Historical Linguistics, 2013

Faster Optimal Planning with Partial-Order Pruning [bib][brief][pdf]
David Hall and Aloni Cohen and David Burkett and and Dan Klein
ICAPS, 2013

Automated reconstruction of ancient languages using probabilistic models of sound change [bib][brief][pdf][press 1, 2, 3, 4][BBC World Service segment]
Alexandre Bouchard-Côté, David Hall, Thomas L. Griffiths, and Dan Klein
Proceedings of the National Academy of Sciences, 2013

Training Factored PCFGs with Expectation Propagation [bib][brief][pdf]
David Hall and Dan Klein
Empirical Methods in Natural Language Processing, 2012.
Distinguished Paper

Parser Showdown at the Wall Street Corral: An Empirical Investigation of Error Types in Parser Output [bib][brief][pdf]
Jonathan K. Kummerfeld and David Hall and James R. Curran and Dan Klein
Empirical Methods in Natural Language Processing, 2012

Iterative Monotonically Bounded A* [bib][brief][pdf]
David Burkett, David Hall and Dan Klein
Association for the Advancement of Artificial Intelligence, 2011

Large-Scale Cognate Recovery [bib][brief][pdf]
David Hall and Dan Klein
Empirical Methods in Natural Language Processing, 2011

Finding Cognate Groups Using Phylogenies [bib][brief][pdf]
David Hall and Dan Klein
Association for Computational Linguistics, Uppsala 2010.

Labeled LDA: A supervised topic model for credit attribution [pdf]
Daniel Ramage, David Hall, Ramesh Nallapati and Christopher D. Manning.
Empirical Methods in Natural Language Processing, Singapore 2009.

Studying the History of Ideas Using Topic Models [pdf]
David Hall, Dan Jurafsky, and Christopher D. Manning.
Empirical Methods in Natural Language Processing, Honolulu, 2008.

Learning Alignments and Leveraging Natural Logic [pdf]
Nathanael Chambers, Daniel Cer, Trond Grenager, David Hall, Chloe Kiddon, Bill MacCartney, Marie-Catherine de Marneffe, Daniel Ramage, Eric Yeh and Christopher D. Manning.
ACL Workshop on Textual Entailment and Paraphrase, Prague, 2007.

Undergraduate Thesis

Tracking the Evolution of Science [pdf]
Honors Thesis. (Advisors: Dan Jurafsky and Christopher Manning.) 2008.

Projects

Breeze
A set of tools for doing NLP, Machine Learning, and whatever else entertains me, in the lovely Scala programming language.

Historical Linguistics
We're working on automatic discovery of cognate groups and the reconstruction of ancient word forms. We're also looking into reconstruction of semantics, morphology, and maybe even syntax.

Overmind
A project to build an agent for playing the game StarCraft. Our goals for this project include the creation of a robust, scalable system that can emulate many different human-like styles of play at different skill levels.

Tolstoy at the Limits
A never-ending project (with Folahan Olowoyeye) to derive a precise understanding of what is meant by Tolstoy's calculus of history in War and Peace. Current avenues of exploration include non-parametric Bayesian statistics, measure theory, and elementary calculus.

Awards

Google PhD Fellowship in Natural Language Processing, 2012.

Outstanding Graduate Student Instructor, EECS, 2011.

Outstanding Graduate Student Instructor, Campus, 2011.

Winner, AIIDE Starcraft AI competition, 2010.

NSF Graduate Research Fellowship, 2010.

Firestone Medal for Undergraduate Thesis, 2008.

Teaching

CS194-13: Large Scale Decision Making in Artificial Intelligence
Spring 2011. (GSI) Harder class. Better Reinforcement Learning. Faster Search. SVMs. Planning. Starcraft.

CS188: Artificial Intelligence
Fall 2010. (GSI) Search. Markov Decision Processes. Reinforcement Learning. Bayes Nets. Probabilistic Tracking. PacMan.

CS124: From Language to Information
Winter 2009. (TA) Natural Language Processing. Social Networks. Information Extraction. Genomics.

CS107: Programming Paradigms
Autumn and Spring, 2007 and 2008. (TA) C Memory Management. Scheme. Concurrency. Code Generation. This class is not what is used to be, for better and for worse.

CS92SI: Explorations on OCaml
Spring, 2007. (Course Leader) Basics of the OCaml programming language. Thinking in terms of modules and lambdas.

CS93SI: Modern C++ Techniques
Spring, 2006. (Course Leader) C++ Templates and Factories. Templates and LISP. Exceptions. Templates and more templates.