CSCI 587 - Lecture 10 Transition Networks

• The Top-Down Parse Algorithm Fails by ...
• Tracing Top Down Parses
• Finite Automata: M = (__ , S, __ , F, s₀)
  • Sigma = the input alphabet
  • S - the set of states
  • delta - the transition function delta(s,a) = newstate
  • F - the set of Final or Accepting states
  • s₀ - the start state
• L(M) = the language accepted by a Finite Automata
  L(M) = {strings with characters from the alphabet such that the sequence of transitions specified by that characters in the string leads from the start state to an accepting state}
• Example: 0*1*
• The Chomsky Hierarchy Box 3.1
  • regular grammars
  • context sensitive
  • unrestricted
• NLP in Prolog: nlp, nlp2
• Parsing with Transition Networks
• Finding equivalent grammar for RTNs
• Augmented Transition Networks
• Test 1 Next Tuesday
  1. HW problems
  2. Material since last HW; i.e., parsing, simple prolog
  3. Construct a Parse tree
  4. Give a top-down parse trace using the grammar
  5. Give an uncle rule(s) in prolog
  6. Concordances
  7. Authorship analysis
  8. Encryption
  9. RSA - no calculators required
  10. Compaction: Huffman Code
  11. Linguistics Review
      • word categories
      • tests for word categories
      • open/closed classes of words
      • head and complement
  12. Eliza Rules

Readings

Sections 3.5

Prolog examples in /class/csci587/Prolog : family, nlp, nlp2

Chapter 1 of Programming in Prolog by Clocksin and Mellish.

Reading Room Sumwalt 203