COMS W4705: Natural Language Processing

[Main] | [General Information] | [Problem Sets]



Instructor: Michael Collins
Time & Location: Tues & Thurs 4.10-5.25, 602 Hamilton Hall
Office Hours:Wednesdays 4-5pm, CEPSR (Schapiro) 723

TAs:
Daniel Bauer (office hours Monday 4-5 and Friday 3-4:30 in the SpeechLab, 7LW3 CEPSR (Shapiro) building)
Sara Rosenthal (office hours Tuesday 1.30-2.30, and Thursday from 3-4, in 726 CEPSR)


Announcements:
Lectures:
Date Topic References
9/6 Introduction and Overview
9/8 Language Modeling Notes on language modeling
(Required reading: updated version posted Sep 15th, fixing a typo in the definition of missing mass on pages 11 and 12.).
9/13 Tagging, and Hidden Markov Models Notes on HMMs
(Required reading)
9/15 Tagging, and Hidden Markov Models (continued)
9/20 Parsing, context-free grammars, and probabilistic CFGs Note on PCFGs (required reading)
9/22 Parsing, context-free grammars, and probabilistic CFGs (continued)
9/27 Parsing, context-free grammars, and probabilistic CFGs (continued)
9/29 Lexicalized probabilistic CFGs
10/4 Quiz 1
10/6 Guest lecture by Nizar Habash
10/11 Lexicalized probabilistic CFGs (continued, see previous slides)
10/13 Machine translation part 1 (Note: we didn't cover the final section, on evaluation using BLEU, but I've kept it in the slides in case it's of interest.)
10/18 Machine translation part 2 Note on IBM Models 1 and 2 (required reading)
10/20 Machine translation part 2 (continued)
10/27 Phrase-based translation models Note on phrase-based models (required reading: updated on 30th October to include details of learning a phrase-based lexicon from training examples (sections 1 and 2))
Slides from the tutorial by Philipp Koehn
11/3 Reordering for statistical MT
11/10 Log-linear models Note on log-linear models (required reading).
11/15 Log-linear tagging (MEMMs)
11/17 Global linear models
11/22 Global linear models (continued)
11/29 Global linear models part II
12/1 Global linear models part III,
Word-sense disambiguation
12/6 The Brown word-clustering algorithm
12/8 The EM algorithm for HMMs