30 August 2010

Lecture 1: Welcome to CL

[Thanks to Kathy McKeown, Ellen Riloff and Jimmy Lin for some material and examples.]

Course administrivia: see web page

About This Course

CL versus NLP?
  • doing what linguists do to language, with computers
  • doing what humans  do to language, with computers
  • science versus engineering
Course Goals:
  • How can machines recognize and generate text (+ speech)?
Knowledge Needed

Goal: Discover what the speaker meant!
Morphology: word formation
  • "Stud tires out"
    • is "tires" a noun or verb?
  • Web search: union activities in New York
    • union/unions?  activities/activity?
    • active?  action?  actor?
    • New versus New York
  • kick => kicks, kicked, kicking
  • sit  => sits,  sat,    sitting
  • gorge / gorgeous
  • glass / glasses
  • arm   / army
Syntax: structural relationships between words
  • Parts of speech
    • The_D boy_N threw_V a_D ball_N to_P the_D brown_J dog_N .
    • Time flies like an arrow ?
  • Constituent Structure
    • John hit Bill.
    • Bill was hit by John.
    • Bill hit John.
    • Bill, John hit.
    • Who John hit was Bill.
  • Structural ambiguity (attachment)
    • Teacher Strikes Idle Kids
    • Enraged Cow Injures Farmer With Ax
    • I saw the Grand Canyon flying to New York.
    • I saw the man on the hill with the telescope.
  • Syntax doesn't tell us much about meaning
    • Colorless green ideas sleep furiously
    • fire match arson hotel
    • "dog collar" versus "flea collar"
    • plastic cat food can cover
Semantics: word meaning and word composition
  • Lexical semantics
    • I walked to the BANK ...
    •              ... of the river.
    •              ... to get money.
    • The BUG in the room ...
    •              ... flew out the window.
    •              ... was planted by spies.
    • I work for John Hancock ...
    •              ... and he is a good boss.
    •              ... which is a good company.
  • Scoping issues
    • Everyone on the island speaks two languages
    • Two languages are spoken by everyone on the island
Discourse: influence of textual context
  • Coreference/anaphoricity:
    •    [President John F. Kennedy] was assassinated.
    •    [The president] was shot yesterday.
    •    Relatives said that [John] was a good father.
    •    [JFK] was the youngest president in history.
    •    [His] family will bury [him] tomorrow.
    •    Friends of [the Massachusetts native] will hold a candlelight service in [Mr. Kennedy]'s home town.
  • Adding interpretations:
    • 1. I only like traveling to Europe.
    •     So I submitted a paper to ACL.
    • 2. I only like traveling to Europe.
    •     Nevertheless, I submitted a paper to ACL.
Pragmatics: influence of context/situation, how we use language to communicate
  •   "Going Home" - A Play in One Act
  Scene 1: Pennsylvania Station, NY
             (Hal) Long Beach?
    (Passerby) Downstairs, LIRR Station.

  Scene 2: Ticket Counter, LIRR Station
             (Hal) Long Beach?
          (Clerk) $4.50.

  Scene 3: Information Booth, LIRR Station
             (Hal) Long Beach?
          (Clerk) 4:19, Track 17.

  Scene 4: On the train, vicinity of Forest Hills
             (Hal) Long Beach?
  (Conductor) Change at Jamaica.

  Scene 5: Next train, vicinity of Lynbrook
             (Hal) Long Beach?
  (Conductor) Right after Island Park.
  • Rules of Conversation:
    • Do you have the time?
    • Could I please have the salt?
  • Speech Acts
    • I bet you $50 that the Nats will win.
    • Will you marry me?
World Knowledge: facts about the world at large, common sense
  • John went to the diner.
       He ordered steak.
       He left a tip and went home.
  • John wanted to commit suicide.
       He got a rope.

Welcome to CMSC 723!

Please use the 'comments' section to ask questions about lectures, homeworks and projects!  Remember to subscribe to this blog's RSS feed!