Course assignment scripts for Natural Language Processing course at the University of Michigan.
Includes the following repositories:
####Language Model Develops and evaluates a language model using a Hidden Markov Model, and uses this language model to train and evaluate a part-of-speech tagger.
####Machine Translation Includes implementation of IBM Models 1 and 2 and the Berkeley Aligner to build a German-English translation system using the Comtrans corpus.
####Multilingual Dependency Parsing Implements a version of Joakim Nivre's acr-eager transition-based dependency parser in English, Danish, and Swedish.
####Word Sense Disambiguation Creates, trains, and evaluates a WSD system in English, Spanish, and Catalan using the SENSEVAL corpus and based on the vector space model.