Computational Linguistics Syllabus |
Course Outline |
Linguistics 581 |
Day |
Reading |
Assignment |
Lecture |
Background |
Code |
Mon Jan 27 | Chapter 1 and Section 2.1 of Chapter 2. Jurafsky and Martin (J&M) History of Computational Linguistics. Regular Expressions and Unix demo. | Assignment 1. | Some history. Textbook intro | What is Computational Linguistics? | |
Mon Feb 03 | Chapter 2, 2.1-2.4 Finite-State Automata | Assignment 2 Installing Python. | Textbook Ch. 2 slides Non-deterministic automata, epsilon transitions | ||
Mon Feb 10 | Reading: Sections 3.1-3.3 of Chapter 3. Introducing words and word parts. Sections 3.4-3.9 of Chapter 3. | Selected exercises from Ch. 3 | Textbook Ch. 3 slides Introduction to transducers: Relations on strings Introduction to transducers | ||
Mon Feb 17 | Section 3.10 of Chapter 3. Spelling Correction. Introduction to Viterbi algorithm. Chapter 4: J&M. 4.1-4.3. Word counting, frequency dictionaries, simple ngram models, the training corpus | Spelling correction assignment | Spelling correction, Brief probability intro | ||
Mon Feb 24 | Chapter 4: J&M (ctd) Ngrams. Introduction to NLTK | NLTK assignment, using Pylab, importing NLTK corpora. | Lecture, Entropy and Cross-Entropy. Entropy as expected information value; cross-entropy as an evaluation tool (pdf, ps) | ||
Mon Mar 03 | Section 4.1-4.5 of chapter 4. Practicalities. Sections 4.5.3 Chapter 4. Sections 4.4 and 4.5.1, 4.5.2 Chapter 4. Smoothing, Add-1 smoothing, Kneser-Ney smoothing. Unknown words. | The Pollard assignment and smoothing assignment are due next week. | Lecture. Smoothing Lecture. Kneser-Ney Lecture. | ||
Mon Mar 10 | 5.1-5.4. Word-class and part of speech tagging. Rule-based taggers. | Tagging Assignment. | Lecture. Tagging slides | ||
Mon Mar 17 | Chapter 5 of the NLTK book.Taggers used on data. | This NLTK-based tagging assignment is due next week. | Tagging slides | ||
Mon Mar 24 | Chapter 5 of the NLTK book.Taggers used on data. | Computing Viterbi by hand. | Tagging slides | ||
Mon Mar 31 | H'day | H'day | H'day | H'day | H'day |
Mon Apr 07 | 6.6 Maximum entropy models | Max entropy assignment, Partial Viterbi answer. | Lecture:: BBoard slides (Slp05.pdf) | ||
Mon Apr 14 | 6.6 Maximum entropy models. Background. Linear regression and logistic regression. 6.7-6.8 Max Ent tagging and Max Ent Markov Models. | Lecture | |||
Mon Apr 21 | Chapter 12. Context Free Grammars of English, Treebanks | Assignment: Midterm, part I Midterm, part II (tagging) NLTK taggiong assignment answer | topdown lecture Lecture: Top down parsing. Parsing as search. | td_parser-0.1:an implementation of a recursive descent top down recognizer. | |
Mon Apr 28 | Chapter 13.1 Parsing. 13.4.1.2 CKY algorithm (bottum up parsing with a chart), Earley algorithm (top down parsing with chart) | Grammar assignment Parsing assignment | |||
Mon May 05 | Chapter 14. Probabilistic Context Free Grammars | Assignment: Prob parsing assignment, CKy implementation in these notes | Lecture | ||
Mon May 12 | Chapter 19. Lexical Semantics. Chapter 20. Computational Lexical Semantics. | Assignment: Lexical semantics assignment | |||
Mon May 19 | Chapter 20. Computational Lexical Semantics. | Assignment: Viterbi problem, Earley algorithm problem, Other problems, Review stuff, Example earley problems with solutions. |