CS626-449: Speech, NLP and the Web/Topics in AI - PowerPoint PPT Presentation

cs626 449 speech nlp and the web topics in ai n.
Download
Skip this Video
Loading SlideShow in 5 Seconds..
CS626-449: Speech, NLP and the Web/Topics in AI PowerPoint Presentation
Download Presentation
CS626-449: Speech, NLP and the Web/Topics in AI

play fullscreen
1 / 86
CS626-449: Speech, NLP and the Web/Topics in AI
192 Views
Download Presentation
felicia-mcdonald
Download Presentation

CS626-449: Speech, NLP and the Web/Topics in AI

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. CS626-449: Speech, NLP and the Web/Topics in AI Pushpak Bhattacharyya CSE Dept., IIT Bombay Lecture 13: Deeper Adjective and PP Structure; Structural Ambiguity and Parsing

  2. Types of Grammar • Prescriptive Grammar • Taught in schools • Emphasis is on usage • Descriptive Grammar • Also known as Linguistic Grammar • Describes Language

  3. Types of Languages • SVO • Subject – Verb – Object • English • E.g. Ramlikesmusic. • S V O • SOV • Subject- Object-Verb • Indian Languages • E.g. रामपानीपी रहा है| • S O V

  4. More deeply embedded structure NP N’1 The AP N’2 N’3 big PP N book with the blue cover PP of poems

  5. Bar-level projections • Add intermediate structures • NP (D) N’ • N’ (AP) N’ | N’ (PP) | N (PP) • () indicates optionality

  6. New rules produce this tree NP N-bar N’1 The AP N’2 N’3 big PP N book with the blue cover PP of poems

  7. As opposed to this tree NP PP The AP PP book with the blue cover of poems big

  8. V-bar • What is the element in verbs corresponding to one-replacement for nouns • do-so or did-so

  9. I [eat beans with a fork] VP PP eat NP with a fork beans No constituent that groups together V and NP and excludes PP

  10. Need for intermediate constituents • I [eat beans] with a fork but Ram [does so] with a spoon VP V1’ VPV’ V’ V’ (PP) V’ V (NP) V2’ PP V NP with a fork eat beans

  11. How to target V1’ • I [eat beans with a fork], and Ram [does so] too. VP V1’ VPV’ V’ V’ (PP) V’ V (NP) V2’ PP V NP with a fork eat beans

  12. Case of conjunction VP V1’ V2’ PP V3’ V4’ Conj and In the afternoon NP NP V V eat drink beans coffee

  13. A-bar: adjectives AP A1’ AP A’ A’ (AP) A’ A’ A (PP) A2’ AP A3’ A5’ Conj and Very A6’ A4’ AP AP green blue bright dull

  14. So-replacement for adjectives • Ram is very serious about studies , but less so than Shyam

  15. P-bar: prepositions AP A1’ PP P’ P’ P’ (PP) P’ P (NP) P1’ PP AP P2’ P3’ Conj and right NP NP P P into off the trash the table

  16. So-replacement for Prepositions • Ram is utterly in debt, but Shyam is only partly so.

  17. Complements and Adjuncts orArguments and Adjuncts

  18. Rules in bar notation: Noun • NP (D) N’ • N’ (AP) N’ • N’ N’ (PP) • N’ N (PP)

  19. Rules in bar notation: Verb • VP V’ • V’ V’ (PP) • V’ V (NP)

  20. Rules in bar notation: Adjective • AP A’ • A’ (AP) A’ • A’ A (PP)

  21. Rules in bar notation: Preposition • PP P’ • P’ P’ (PP) • P’ P (NP)

  22. Introducing the “X factor” • Let X stand for any category N, V, A, P • Let XP stand for NP, VP, AP and PP • Let X’ stand for N’, V’, A’ and P’

  23. XP to X’ • Collect the first level rules • NP (D) N’ • VP V’ • AP A’ • PP P’ • And produce • XP (YP) X’

  24. X’ to X’ • Collect the 2nd level rules • N’ (AP) N’ or N’ (PP) • V’ V’ (PP) • A’ (AP) A’ • P’ P’ (PP) • And produce • X’ (ZP) X’ or X (ZP)

  25. X’ to X • Collect the 3rd level rules • N’ N (PP) • V’ V (NP) • A’ A (PP) • P’ P (NP) • And produce • X’ X (WP)

  26. Basic observations about X and X’ • X’ X (WP) • X’ X’ (ZP) • X is called Head • Phrases must have Heads: Headedness property • Category of XP and X must match: Endocentricity

  27. Basic observations about X and X’ • X’ X (WP) • X’ X’ (ZP) • Sisters of X are complements • Roughly correspond to objects • Sisters of X’ are Adjuncts • PPs and Adjectives are typical adjuncts • We have adjunct rules and complement rules

  28. Structural difference between complements and adjuncts XP X’ ZP X’ Adjunct WP X Complement

  29. Complements and Adjuncts in NPs NP N’ ZP N’ PP with red cover N book of poems

  30. Any number of Adjuncts NP N’ N’ ZP from Oxford Press N’ PP with red cover N book of poems

  31. Parsing Algorithm

  32. A simplified grammar • S  NP VP • NP  DT N | N • VP  V ADV | V

  33. A segment of English Grammar • S’(C) S • S{NP/S’} VP • VP(AP+) (VAUX) V (AP+) ({NP/S’}) (AP+) (PP+) (AP+) • NP(D) (AP+) N (PP+) • PPP NP • AP(AP) A

  34. Example Sentence People laugh • 2 3 Lexicon: People - N, V Laugh - N, V These are positions This indicate that both Noun and Verb is possible for the word “People”

  35. Top-Down Parsing State Backup State Action ----------------------------------------------------------------------------------------------------- 1. ((S) 1) - - 2. ((NP VP)1) - - 3a. ((DT N VP)1) ((N VP) 1) - 3b. ((N VP)1) - - 4. ((VP)2) - Consume “People” 5a. ((V ADV)2) ((V)2) - 6. ((ADV)3) ((V)2) Consume “laugh” 5b. ((V)2) - - 6. ((.)3) - Consume “laugh” Termination Condition : All inputs over. No symbols remaining. Note: Input symbols can be pushed back. Position of input pointer

  36. Discussion for Top-Down Parsing • This kind of searching is goal driven. • Gives importance to textual precedence (rule precedence). • No regard for data, a priori (useless expansions made).

  37. Bottom-Up Parsing Some conventions: N12 S1? -> NP12 ° VP2? Represents positions Work on the LHS done, while the work on RHS remaining End position unknown

  38. Bottom-Up Parsing (pictorial representation) S -> NP12 VP23° People Laugh 1 2 3 N12 N23 V12 V23 NP12 -> N12 ° NP23 -> N23 ° VP12 -> V12 ° VP23 -> V23 ° S1? -> NP12° VP2?

  39. Problem with Top-Down Parsing • Left Recursion • Suppose you have A-> AB rule. Then we will have the expansion as follows: • ((A)K) -> ((AB)K) -> ((ABB)K) ……..

  40. Combining top-down and bottom-up strategies

  41. Top-Down Bottom-Up Chart Parsing • Combines advantages of top-down & bottom-up parsing. • Does not work in case of left recursion. • e.g. – “People laugh” • People – noun, verb • Laugh – noun, verb • Grammar – S  NP VP NP  DT N | N VP  V ADV | V

  42. Transitive Closure People laugh 1 2 3 S NP VP NP N VP  V  NP DT N S  NPVP S  NP VP  NP N VP V ADV success VP V

  43. Arcs in Parsing • Each arc represents a chart which records • Completed work (left of ) • Expected work (right of )

  44. Example People laugh loudly 1 2 3 4 S  NP VP NP  N VP  V VP  V ADV NP  DT N S  NPVP VP  VADV S  NP VP NP  N VP V ADV S  NP VP VP V

  45. Dealing With Structural Ambiguity • Multiple parses for a sentence • The man saw the boy with a telescope. • The man saw the mountain with a telescope. • The man saw the boy with the ponytail. At the level of syntax, all these sentences are ambiguous. But semantics can disambiguate 2nd & 3rd sentence.

  46. Prepositional Phrase (PP) Attachment Problem V – NP1 – P – NP2 (Here P means preposition) NP2 attaches to NP1 ? or NP2 attaches to V ?

  47. Parse Trees for a Structurally Ambiguous Sentence Let the grammar be – S  NP VP NP  DT N | DT N PP PP  P NP VP  V NP PP | V NP For the sentence, “I saw a boy with a telescope”

  48. Parse Tree - 1 S NP VP N V NP Det N PP saw I a P NP boy Det N with a telescope

  49. Parse Tree -2 S NP VP PP N V NP P NP Det N saw I Det N with a boy a telescope

  50. Parsing Structural Ambiguity