Computational linguistics allows computers to use language, such as recognizing speech, correcting spelling, and translating. This course introduces students to the field using a modern statistical approach.
Week | Date | Topic | Reading | Materials | Assignments |
---|---|---|---|---|---|
1.1 | Mar 28 | What is computational linguistics, unix/linux | JM 1, unix/linux tutorial (through tutorial 5) | Connect to the SSCC [Instructions] | |
1.2 | Mar 30 | No class. Klinton traveling. | |||
2.1 | Apr 4 | Programming in python 1 | NLTK 1 | python transcript | |
2.2 | Apr 6 | Finite-state automata, regular expressions | JM 2 | nltk setup, nano tutorial, transcript, optional: emacs tutorial | hw1 out |
3.1 | Apr 11 | Programming in python 2 | NLTK 2–3 | ||
3.2 | Apr 13 | Programming in python 3 | NLTK 4 | lecture notes | |
4.1 | Apr 18 | Probability theory, maximum likelihood estimation (MLE), unigram models | hw1 due | ||
4.2 | Apr 20 | Graphical models, n-gram models, Markov chains | JM 4.1–4.2, Levy appendix | ||
5.1 | Apr 25 | Perplexity, training and test sets, basic information theory | JM 4.3–4.4 | hw2 out | |
5.2 | Apr 27 | More smoothing, part-of-speech tagging | JM 4.5–4.7, 4.9.1, 4.10 | ||
6.1 | May 2 | Bayesian inference, Hidden Markov models (HMMs), part-of-speech tagging | JM 5.1–5.3, 5.5–5.5.2, 5.7 | hw2 due; hw3 out | |
6.2 | May 4 | Forward algorithm | JM 5.5.3, 6.1–6.4 | ||
7.1 | May 9 | Viterbi decoding; Programming: best practices | Wilson et al. (2014) | scripts archive, slides | |
7.2 | May 11 | Supervised and unsupervised learning, noisy channel models for spelling correction/autocorrect | JM 5.5.4, JM 5.9 | ||
8.1 | May 16 | Context-free grammars (CFGs) for syntax, classes of grammars, regular expressions on trees, basic parsing | JM 12.1–12.6, 13.1–13.4.2 | hw3 due; project proposals due; hw4 out | |
8.2 | May 18 | Probabilistic CFGs (PCFGs), statistical parsing | JM 14.1–14.4 | ||
9.1 | May 23 | Automatic speech recognition (ASR), Machine translation (MT) | JM 9.1–9.2, 9.5–9.6; 25.1–25.3 | ||
9.2 | May 25 | Computational psycholinguistics | Bicknell & Levy (2010) | hw4 due | |
Jun 5 | Final project reports due 5pm |