Ling 361, Intro to Computational Linguistics
fall 2002 syllabus

General course information

Course Description

This course presents an overview of the field of computational linguistics with an emphasis on its applications. Some major applications include information retrieval, question-answering systems, and machine translation; we will touch on others also.

Though no programming is required, an acquaintance with programming, algorithms, mathematics, statistical methods, or logic will be helpful at various points. Some familiarity with linguistics is assumed as well. Students will complete a small project, and there will be homework assignments, a midterm exam, and a final exam.

Instructor

• Dr. Tony Davis -- ard8@georgetown.edu or
  tdavis@onebox.com office hours: Mondays 10:30 - 11:30 and Wednesdays 1:00 - 2:00, in ICC 456

Lectures

• Class meets on Mondays and Wednesdays from 11:40 to 12:55 in ICC, room 117
• Try to do the reading listed for each lecture before class.

Course Requirements

• Evaluation/Grading:

  • Midterm exam - 20%
  • Final exam - 25%
  • Homework assignments - 20%
  • Project/Presentation - 25%
  • Class participation - 10%
  
  

• Homework assignments

  There will be 5 or 6 homework assignments throughout the semester. Collaborating on these with others in the class is encouraged. However, each student should turn in an assignment individually, and indicate who (s)he is collaborating with. Assignments are due at the end of class on the indicated day. You may turn in one assignment by the end of the following class without penalty; subsequent late assignments will be marked down.
  



• Exams

  Exams will be "closed book". The midterm exam is on Wed., Oct. 16 (week 8), and will last about an hour. The final exam is on Thursday, December 12, in ??, from 4 to 6 pm. Make-up exams will not be given except for valid medical or other emergency reasons.
  



• Project/Presentation

  See the project requirements and ideas page for more information.

  Each student will complete a small project drawing on the themes and techniques we will study in the course. The project might involve evaluating or comparing existing tools (for example, machine translation tools available on the web or document categorization systems), a research paper extending beyond the material covered in class (for example, methods of probabilistic parsing or named entity extraction), or programming and implementing a small system.

  I will meet with each of you midway through the course to discuss your project. Towards the end of the course, everyone will hear you give a presentation of your project (probably about 15 or 20 minutes long).
  




• Class participation

  Raising questions and points for discussion in class is strongly encouraged. Some of the readings are primary sources and it should be helpful to work out issues, problems, and unclear points together. Try to read the assigned material for each session beforehand, and come with questions and reactions.

Readings

• Textbooks

  • We'll be reading several chapters from Speech and Language Processing by Daniel Jurafsky and James H. Martin. The website for the book offers some links for ordering it, a list of errata, and a downloadable pdf file of chapter 1, so that you can get started with this before buying the book.
  • We'll also draw on Foundations of Statistical Language Processing by Christopher D. Manning and Hinrich Sch|tze. This book also has a companion website with errata, ordering information, and a couple of sample chapters. Chapter 1 and a portion of 15 are in the reader, and there's a downloadable version of chapter 5, which we will also be covering.
  
  

• Reader

  The reader will be available by Wednesday, Aug. 28, in class; you can purchase it directly from me for $6.25. It contains portions of two chapters from Manning and Sch|tze's book, excerpts from chapter 1 of Terry Winograd's book Language as a Cognitive Process and some papers that aren't available for downloading from the internet.

• Papers on the web

  These can be downloaded and printed at your convenience (but don't wait until the last minute!). Please let me know if you have trouble obtaining any of them.

Additional resources

• The websites for the Jurafsky and Martin and Manning and Sch|tze books have a lot of useful links.
  For example, this link to HyperStat Online
has an introduction to probability with a lot of examples, and links to other, similar sites.
• Similarly, many of the authors whose websites you will download readings from have interesting links.
• Some parsers you can try online:
  • Conexor's analyzers: http://www.conexoroy.com/products.htm
  • The CMU Link Parser: http://www.link.cs.cmu.edu/link/
  • Some sample output from these parsers
• WordNet WordNet home page
  another WordNet interface
• The Association for Computational Linguistics publishes the quarterly journal Computational Linguistics and organizes some of the leading conferences in the field.
• The Nitrogen text generation system at ISI.

Week by Week

This schedule is tentative and will certainly change, so check frequently for updates.

Week 1	Wed., Aug. 28	What is Computational Linguistics? What makes CL hard?	Winograd, ch. 1, and Manning and Sch|tze, ch. 1
Week 2	Mon., Sept. 2	Labor Day Holiday (no class, but continue with reading)	Jurafsky and Martin, ch. 1 and look at HyperStat Online (an introduction to probability), particularly the links in the Contents column
	Wed., Sept. 4	Applications of CL Symbolic and statistical approaches to CL Some basics of probability and statistics	Jurafsky and Martin, ch. 2 first homework assignment
Week 3	Mon., Sept. 9	Regular Extressions and Finite-State Automata Morphology and Finite State Transducers	Jurafsky and Martin, ch. 3
	Wed., Sept. 11	Information Retrieval: Introduction	Manning and Sch|tze, ch. 15, sec. 1 & 2 (portion in reader) Jurafsky and Martin, ch. 17, sec. 3 interpolated homework assignment (Jurafsky and Martin, problems 2.4 and 2.8)
Week 4	Mon., Sept. 16	Information Retrieval: the vector space model Search Engines by Tim Sibley "Precision Content Retrieval"	Ambroziak and Woods (1998), Flank (2000) first homework assignment due
	Wed., Sept. 18	N-Grams and Collocations	Jurafsky and Martin, ch. 6, sec. 1 & 2 (pp.191-206) Manning and Sch|tze, ch. 5, through section 5.2 second homework assignment interpolated homework assignment due
Week 5	Mon., Sept. 23	Collocations (continued)	Manning and Sch|tze, ch. 5, section 5.3 to end
	Wed., Sept. 25	Part of Speech Tagging	Jurafsky and Martin, ch. 8
Week 6	Mon., Sept. 30	Context-free Grammars and Parsing	Jurafsky and Martin, ch. 10 (also look over ch. 9 if you aren't familiar with context-free grammars) second homework assignment due
	Wed., Oct. 2	Question-Answering Sytems	Cardie et al. (2000), Molla Aliod, Berri, and Hess (1998)
Week 7	Mon., Oct. 7	Lexical Ambiguity Resolution	Resnik and Yarowsky (1997) or pdf here Pedersen (2001)
	Wed., Oct. 9	more on Lexical Ambiguity Resolution Midterm Review	(continued from previous lecture)
Week 8	Mon., Oct. 14	Columbus Day Holiday -- no class	review for midterm exam
	Wed., Oct. 16	Midterm exam	review for midterm exam
Week 9	Mon., Oct. 21	Syntactic Ambiguity Resolution	Hindle and Rooth (1993) Whittemore, Ferrara, and Brunner (1990)
	Wed., Oct. 23	Semantics and Interpretation: Lexical Semantics Barrett, Davis, and Dorr PowerPoint slides	Barrett, Davis, and Dorr (2001), Richardson, Dolan, and Vanderwende (1998) third homework assignment
Week 10	Mon., Oct. 28	Referent Resolution	Jurafsky and Martin, ch. 18
	Wed., Oct. 30	Discourse and Dialog Systems	Jurafsky and Martin, ch. 19 third homework assignment due
Week 11	Mon., Nov. 4	Document Summarization	Hahn and Mani (2000), Buyukkokten, Garcia-Molina, and Paepcke (2001) project proposal due
	Wed., Nov. 6	Generating Natural Language Demo of the Nitrogen NLG system from ISI	Jurafsky and Martin, ch. 20
Week 12	Mon., Nov. 11	Machine Translation Bonnie Dorr's slides	Jurafsky and Martin, ch. 21 final homework assignment: exercises 18.11, 19.4, and 21.5 in Jurafsky and Martin
	Wed., Nov. 13	Machine Translation (reprise) See MultiMeteo's weather forcasts in several languages	Coch and Chevreau (2001) the MultiMeteo website
Week 13	Mon., Nov. 18	Putting it all together: The Verbmobil project;	Malouf and Riehemann's slides
	Wed., Nov. 20	Multilingual Information Retrival	Flank (2000b) work on projects
Week 14	Mon., Nov. 25	StreamSage demo	work on projects
	Wed., Nov. 27	Class Presentations	work on projects final homework assignment due
Week 15	Mon., Dec. 2	Class Presentations; Review	(these are optional) Multilingual Information Management: Current Trends and Future Abilities (the introduction, ch. 1 (you can skip sec. 2), ch. 2, and ch. 6, sec. 4 are probably most interesting) Nunberg (2000)
	Wed., Dec. 4	Wrap-up and Review: The Future of CL	review for final exam project write-ups due

Final exam: Thursday, December 12, 4 - 6 pm in ??.