This schedule is tentative and subject to change as the semester progresses.
Date | Lecture | Reading | Assignment | ||
---|---|---|---|---|---|
Introduction | |||||
W | 8/31 | 1 | What is NLP? | SLP2 Ch. 1 | |
Th-F | 9/1-2 | One-on-one meetings with instructor | |||
M | 9/5 | No class: Labor Day | |||
W | 9/7 | 2 | Why do we need data?; Working with text in Python 3 | NLTK Book 3.2–3.10 | |
Sa | 9/10 | A0: Text processing in Python and Unix | |||
M | 9/12 | 3 | Working with Unix: text processing; version control with git | Unix Text Commands, Git Commands | |
W | 9/14 | 4 | Overview of Linguistics | SLP2 1.1, 3.1, 25.1 | |
F | 9/16 | A0 due | |||
Words & BoW: Supervised | |||||
M | 9/19 | 5 | Classification: naïve Bayes | SLP3 7.0–7.1; Probability review: Goldwater; Manning & Schütze Ch. 2 | |
W | 9/21 | 6 | More smoothing, tuning, and evaluation; Lexical semantics: senses, relations, classes | SLP3 7.2–7.3, 18.0–18.5 | |
M | 9/26 | 7 | Linear models for classification: features & weights | SLP3 7.4.1 | |
Tu | 9/27 | A1: NLI: Classifiers for Native Language Identification | |||
W | 9/28 | 8 | Linear models for classification: discriminative learning (perceptron, SVMs, MaxEnt) | SLP3 7.4; Daumé The Perceptron;
Eisenstein Notes, Ch. 2: Discriminative Learning (Further readings are suggested in slides) |
|
N-grams & Sequences: Supervised | |||||
M | 10/3 | 9 | N-gram language models | SLP3 4.0–4.4 | |
W | 10/5 | 10 | More LM smoothing, noisy channel model; Parts of speech | SLP3 6.0–6.2, 9.0–9.1 | |
M | 10/10 | No class or office hours: Columbus Day | A1 due | ||
W | 10/12 | 11 | POS tagging: HMMs | SLP3 8.0–8.2, 9.3–9.4.2; Eisenstein Notes, Ch. 8: Part-of-speech tagging | |
F | 10/14 | Hal Daumé (UMD): Learning Language through Interaction | CS Colloquium | St. Mary’s 326, 11:00 | |
M | 10/17 | 12 | Algorithms for HMMs (mainly Viterbi) | SLP3 8.4, 9.4.3 | |
MIDTERM REVIEW (James, 5:00) | |||||
W | 10/19 | 13 | MIDTERM EXAM | Study guide | |
M | 10/24 | 14 | Discriminative tagging with the structured perceptron | Eisenstein Notes, 9.3–9.4 (no need to read beyond structured perceptron); Neubig slides | |
W | 10/26 | 15 | Annotation; SPECIAL ACTIVITY | A2: annotation (involves group work) | |
Hierarchical Sentence Structure | |||||
M | 10/31 | 16 | English syntax, CFGs | SLP2 12.1–12.4 | |
W | 11/2 | 17 | Information Retrieval (guest lecture by Grace Hui Yang) | SLP2 23.1 | |
F | 11/4 | A2 due | |||
M | 11/7 | 18 | (P)CFG parsing | SLP2 13.4–13.4.1, 14.1–14.6.0, 14.7 | |
W | 11/9 | 19 | (P)CFGs contd.; Dependency parsing | SLP2 12.4.4, 12.7 | P1: 1-2 page proposal due |
F | 11/11 | Yulia Tsvetkov (CMU/Stanford): On the Synergy of Linguistics and Machine Learning in Natural Language Processing | Linguistics Speaker Series | Poulton 320, 3:30 | |
M | 11/14 | 20 | Dependency parsing contd. | ||
W | 11/16 | 21 | Semantic role labeling | SLP3 22.0–22.6 | |
W-Th | 11/16-17 | P2: groups meet with instructor & TA | |||
F | 11/18 | P3: Progress update, including lit review, due | |||
Marine Carpuat (UMD): Toward Natural Language Inference Across Languages | Linguistics Speaker Series | Poulton 320, 3:30 | |||
Unsupervised Learning | |||||
M | 11/21 | 22 | Machine translation | SLP2 Ch. 25 | |
Shomir Wilson (UC): Text Analysis to Support the Privacy of Internet Users | CS Colloquium | St. Mary’s 326, 11:00 | |||
W | 11/23 | CLASS CANCELED FOR THANKSGIVING | |||
Th-F | 11/24-25 | Thanksgiving Break | |||
M | 11/28 | 23 | More MT; Word & document representations; Distributional similarity | SLP3 Ch. 15, 16.0–16.1, 16.4 | |
Tu | 11/29 | Mark Dredze (JHU): Topic Models for Identifying Public Health Trends | CS Colloquium | St. Mary’s 326, 11:00 | |
W | 11/30 | 24 | Neural networks (James) | Recommended by James: Nielsen Neural Networks and Deep Learning, Cristopher Olah's blog, Rojas Neural Networks, Goodfellow et al. Deep Learning. Also: Dyer et al. Practical Neural Networks for NLP | |
Th | 12/1 | P4: PROJECTS DUE @ 11:59pm | |||
F | 12/2 | Mona Diab (GW): Processing Arabic Social Media: Challenges and Opportunities | CS Colloquium | St. Mary’s 414, 2:30 | |
M | 12/5 | 25 | PROJECT PRESENTATIONS I | ||
W | 12/7 | 26 | PROJECT PRESENTATION II + Context in language processing; Wrap-up | ||
Tu | 12/13 | 4:00-6:00pm: FINAL EXAM, usual room | Study guide | ||
Visit the GUCL website for NLP talks in Spring 2017 and beyond! |