Schedule
- Week 1: Aug 29
- Week 2: Sept 3, 5
- Week 3: Sept 10, 12
- Week 4: Sept 17, 19
- Week 5: Sept 24, 26
- Week 6: Oct 1, 3
- Week 7: Oct 8, 10
- Week 8: Oct 15, 17
- Week 9: Oct 22, 24
- Week 10: Oct 29, 31
- Week 11: Nov 5, 7
- Week 12: Nov 12, 14
- Week 13: Nov 19, 21
- Week 14: Nov 26, 28
- Week 15: Dec 3, 5
This is a tentative schedule of the course. All dates are subject to change based on how the class progresses.
Week 1: Aug 29
Thursday, Aug 29
Topics
- Course Introduction
- Introduction to Scala (1/3)
Materials
Assignments- Assignment 0: Programming in Scala, Assigned
Week 2: Sept 3, 5
Tuesday, Sept 3
Topics
- Scala (2/3)
Thursday, Sept 5
Topics
- Scala Questions Day (3/3)
Materials
- Your own questions!
Week 3: Sept 10, 12
Tuesday, Sept 10
Topics
- Probability
Materials
- Probability overview slides (by Markus Dickinson)
- Univariate Probability, chapter from Roger Levy’s draft book Probabilistic Models in the Study of Language.
- Probability Theory Review from Andrew Ng’s Machine Learning course
- Assignment 0: Programming in Scala, Due (by noon!)
- Assignment 1: Probability, Assigned
Thursday, Sept 12
Topics
- Text classification (1/3)
- Naive Bayes
Materials
Week 4: Sept 17, 19
Tuesday, Sept 17
Topics
- Instructor solutions to Assignment 0
- (Add-Lambda) Smoothing for Naive Bayes (2/3)
Thursday, Sept 19
Topics
- Train/Dev/Test Datasets
- Precision and Recall
- Semi-supervised learning for Naive Bayes with Gibbs Sampling (3/3)
- Assignment 1: Probability, Due (programming submitted by noon, written by 2pm)
- Assignment 2: Text Classification, Assigned
Week 5: Sept 24, 26
Tuesday, Sept 24
Topics
- Language Modeling
- N-Gram Language Models (1/4)
Materials
- JM 4: 83-97
- JM 4: 98-113
- Language modeling slides
- Language modeling slides (by Ray Mooney)
Thursday, Sept 26
Topics
- N-Gram Language Models (2/4)
- Maximum Likelihood Estimation (MLE) for N-Gram Models
Week 6: Oct 1, 3
Tuesday, Oct 1
Topics
- N-Gram Language Models (3/4)
- Add-Lambda Smoothing
- Good-Turing, Knesser Ney
- Assignment 2: Text Classification, Due (programming submitted by noon, written by 2pm)
Thursday, Oct 3
Topics
- N-Gram Language Models (4/4)
- Interpolation, Backoff
- Perplexity
- Assignment 3: N-Gram Language Models, Assigned
Week 7: Oct 8, 10
Tuesday, Oct 8
Topics
- Part-of-Speech Tagging
- Hidden Markov Models (1/4)
Materials
- JM 5: 123-163
- POS tagging and HMM slides (by Ray Mooney)
Thursday, Oct 10
Topics
- Hidden Markov Models (2/4)
- HMM Assumptions
- MLE
- Add-Lambda Smoothing
- One-Count Smoothing
Materials
- JM 6: 173-192
Week 8: Oct 15, 17
Tuesday, Oct 15
Topics
- Hidden Markov Models (3/4)
- Decoding (Viterbi Algorithm)
- Tag Dictionaries
Materials
- An Interactive Spreadsheet for Teaching the Forward-Backward Algorithm (by Jason Eisner)
- Assignment 3: N-Gram Language Models, Due (programming submitted by noon, written by 2pm)
- Assignment 4: HMMs, Assigned
Thursday, Oct 17
Topics
- Hidden Markov Models (4/4)
- Semi-Supervised Learning (Forward-Backward Algorithm)
- Low-Resource POS-Tagger Learning
Materials
Week 9: Oct 22, 24
Tuesday, Oct 22
Mid-Term ExamThursday, Oct 24
Topics
- Discriminative models
- Logistic Regression
- Maximum Entropy (MaxEnt) models (1/3)
Materials
- JM 6: 193-207
Week 10: Oct 29, 31
Tuesday, Oct 29
Topics
- Maximum Entropy (MaxEnt) models (2/3)
- Maximum Entropy Markov Models
Thursday, Oct 31
Topics
- Maximum Entropy (MaxEnt) models (3/3)
- Information extraction (NER, relation extraction, co-ref, SRL)
Materials
- JM 6: 207-212
- JM 22: 725-743
- Information extraction slides (by Jim Martin, slightly remixed)
- Assignment 4: HMMs, Due (programming submitted by noon, written by 2pm)
- Assignment 5: MaxEnt, Assigned
Week 11: Nov 5, 7
Tuesday, Nov 5
Topics
- Grammar
- Parsing (1/4)
Materials
- JM 12.
- JM 13. 427-443
Thursday, Nov 7
Topics
- Statistical Parsing (2/4)
Materials
- JM 14: 459-467
Week 12: Nov 12, 14
Tuesday, Nov 12
Topics
- Parsing (3/4)
- CKY
Materials
- JM 17: 545-572
Thursday, Nov 14
Topics
- Parsing (4/4)
- Other grammatical formalisms (TAG, CCG, dependency parsing)
Materials
Assignments- Assignment 5: MaxEnt, Due (programming submitted by noon, written by 2pm)
- Assignment 6: Parsing, Assigned
Week 13: Nov 19, 21
Tuesday, Nov 19
Topics
- Computational Semantics
- Logical Semantics
- Compositionality
- Entailment
Materials
- JM 18: 583-591
Thursday, Nov 21
Topics
- Distributional Semantics
- Vector Spaces
Materials
- JM 17: 545-572
Week 14: Nov 26, 28
Tuesday, Nov 26
Topics
- Weakly-Supervised Learning
- Active Learning
Materials
- Learning a Part-of-Speech Tagger from Two Hours of Annotation by Garrette and Baldridge
- Kristen Grauman’s Computers and Thought Award talk
- Jason Baldridge and Alexis Palmer, How well does active learning actually work?
Thursday, Nov 28
No Class - Thanksgiving
Week 15: Dec 3, 5
Tuesday, Dec 3
Topics
- Topic Models
- Course Evaluations
Materials
- Probabilistic topic models (Steyvers and Griffiths 2007)
- David Blei’s topic modeling talk: video and slides.
- Assignment 6: Parsing, Due (programming submitted by noon, written by 2pm)