▌ ▐ ▐▀▌ █ █▀▌ ▐▀▌ ▌ ▐ ▐▌▐▌ ▐▀▀ ▀▌▐▌▐▀ █▀█ █▀█ █ █ █ █▀█ █ █ ▐▐▌█ █▀ ▌▐▌▐ ▌ ▐ ▌ ▐ ▐██ ███ ▌ ▐ █▄█ ▐ █ ▐██ ▄▌▐▌▐▄ (W)elcome (A)dvisees (T)alks (P)ublications (V)ita/Bio/Contact (F)AQ (T)eaching (B)log (L)inks
I have a few pieces of software I've made available:
  • TagChunk: A Joint POS Tagger and Syntactic Chunker

    Software for performing joint inference over part of speech tags and syntactic chunks; based on my ICML 2005 paper, but not an exact replica. Now also features weights for named entity recognition.

  • HBC: Hierarchical Bayes Compiler

    Simulate, or compile and run hierarchical Bayes models. In Haskell.

  • Maximum Entropy Model Optimization Package

    Optimization software for maximum entropy models, uses conjugate gradient for binary/binomial problems and LM-BFGS for multiclass problems; written in O'Caml

  • DPSearch

    Implementation (in matlab) of the search algorithms from Fast search for Dirichlet process mixture models paper.

  • FastDT

    Very fast decision tree learner that implements bagging and boosting; in O'Caml

  • SVMsequel

    Optimization software for support vector machines, very fast, support multiclass classification and probabilistic outputs, written in O'Caml


    A high accuracy parser of name internal structure. This is a simple rule-based perl script for parsing names.

  • O'Caml Suffix Trees

    An implementation of Ukkonnen's suffix tree algorithm in O'Caml. Code is very C like, with lots of pointers and what-not, but it is useful (used in SVMsequel, above).

  • Haskell - All In One

    Software to convert multi-module Haskell programs into a single module

  • HWordNet

    A Haskell interface to the WordNet database.

  • GDC - Generic Data Clustering

    Software to perform unsupervised data clustering.

  • HBlas/HLAPack

    A port of the BLAS (Basic Linear Algebra System) and a partial port of the LAPack (Linear Algebra Package) to Haskell from Fortran through C.

  • APP - The Haskell Array Preprocessor

    Performs simple preprocessing on Haskell-like source code to allow for convenient array update/read notation.