Computational Linguistics, Winter 2005/06

Administration

Instructor:: Shuly Wintner, shuly@cs.haifa.ac.il. Office hours: Wednesday 12:00-13:00, Jacobs 403. Phone: (828)8180.
Teaching assistant:: Yael Cohen-Sygal, yaelc@cs.haifa.ac.il. Office hours: Thursday 13:00-14:00, Education 561. Phone: (828)8459.
Times:: Monday, 9:00-10:00 and Wednesday, 9:00-12:00.
Place:: Jacobs 209, Computer Lab (Mon), Education 462 (Wed).
Prerequisites:: Computational models (203.3510). Students without the required background should talk to the instructor prior to registration.
Attendance:: Optional but highly recommended.
Grading:: The final grade will be based on 4-6 home assignments (approximately 33%) and two mid-term exams (approximately 33% each). There will be no final exam.
Mid-term exams:: The first mid-term will cover the first half of the material and will take place on December 14th (during class time). The second mid-term will cover the other half of the material and will take place on January 25st (during class time).
Home assignments:: Approximately 5 home assignments will be handed out during the semester. Submission is mandatory. Homework is to be done independently and plagiarism will not be tolerated.

Content

Textbook:

There is no recommended textbook for this course, but some of the material can be found in Speech and Language Processing, by Daniel Jurafsky and James H. Martin. It's a good book, albeit too shallow sometimes. See a recommended bibliography list.

Syllabus:

Introduction

Applications of natural language processing
The structure of natural languages
Different levels of language understanding: morphology, syntax and semantics
Ambiguity

Morphology

Introduction to morphology: word structure
Inflections and derivations
Finite-state automata
Finite-state transducers

Syntax

Introduction to syntax: the structure of natural languages
Context-free grammars: grammars, forms, derivations, trees, languages
Parsing: top-down, CYK algorithm, Earley algorithm, bottom-up chart parsing
The complexity of natural language
The limitations of CFGs
Unification grammars: feature structures and unification

Empirical approaches to Natural Language Processing

Probabilistic techniques
Applications

Announcements

January 27th, 2006: Final grades published. Please check and report any discrepancies. Appeals? Call me to set a time if you'd like to see your graded exam.
January 23th, 2006: The exam will take place in the classroom, Education 462.
January 12th, 2006: Submission of Challenge 4 extended to January 22nd.
January 11th, 2006: A unification grammars workshop will take place on Monday, January 16th, 9:00-10:00 in Jacobs 209.
January 1st, 2006: A unification grammars workshop will take place on Monday, January 9th, 9:00-10:00 in Jacobs 209.
December 30th, 2005: Bibliography list added.
December 26th, 2005: A grammars workshop will take place on Monday, January 2nd, 9:00-10:00 in Education 570.
December 22nd, 2005: Challenge 2 grades published.
December 20th, 2005: Due to popular demand I've decided to hold a Moed B for the midterm exam. The exam will take place on Monday, 2.1.06, at 9:00, in the Lab (Jacobs 209). If you wish to take the exam, you must sign up: just send me an e-mail saying you'd like to take it. The grade of the Moed B exam supersedes the original grade.
December 15th, 2005: First midterm exam and grades. Midterm1 column is the grade of problem 1 (0-100), Midterm 2 is of problem 2 (0-10) and Midterm3 is of problem 3 (0-10). The final midterm grade is (P1 * 0.8 + P2 + P3 + 10).
December 12th, 2005: Midterm of 2004 with solution.
December 12th, 2005: The midterm exam will take place in Jacobs 209 (the Lab).
December 7th, 2005: A parsing workshop will take place on Monday, December 19th, 9:00-10:00 in Jacobs 209.
November 30th, 2005: Submit Challenge 2 by Sunday December 11th.
November 30th, 2005: An XFST lab will take place on Monday, December 5th, 9:00-10:00 in Jacobs 209.
November 27st, 2005: Challenge 1 grades published.
November 27st, 2005: XFST tutorial added.
November 21st, 2005: First mid-term exam date set to December 14th, during class.
November 21st, 2005: Challenge 2 added.
November 17th, 2005: An XFST lab will take place on Monday, November 28th, 9:00-10:00 in Jacobs 209.
November 10th, 2005: Challenge 1 added.
November 6th, 2005: First lecture slides added.
October 19th, 2005: Course web site on-line.

Copies of the slides

Slides by topic: introduction; morphology; finite-state technology, part I; finite-state technology, part II; syntax; context-free grammars; parsing; the complexity of natural languages; unification grammars, part I; unification grammars, part II; unification grammars, part III; empirical methods, part I; empirical methods, part II.

Recitation notes: CFGs; PC_PATR.

Home assignments

Challenge 1. Submit by class time on November 23rd.
Challenge 2. Submit by Sunday December 11th.
Challenge 3. Submit by Sunday January 1st.
Challenge 4. Submit by Sunday January 22nd.