1 / 16

Unification Grammars

Unification Grammars. Allen ’ s Chapter 4 J&M ’ s Chapter 11. Unification of Feature Structures. Key concept - extension relationship between two FSs: F2 extends F1 if every feature-value pair in F1 is also in F2

kapila
Télécharger la présentation

Unification Grammars

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Unification Grammars Allen’s Chapter 4 J&M’s Chapter 11

  2. Unification of Feature Structures • Key concept - extension relationship between two FSs: • F2 extends F1 if every feature-value pair in F1 is also in F2 • Two FSs F1 and F2 unify if there exists a FS F that is an extension of both of them • The Most General Unifier is the minimal FS F that extends both

  3. Unification of Feature Structures • F2 extends F1: F1 = (CAT v) F2 = (CAT v ROOT cry) F1 = (AGR {3s 3p}) F2 = (AGR 3s) • F3 is MGU of F1 and F2: F1 = (CAT v ROOT cry), F2 = (CAT v VFORM pres), F3 = (CAT v ROOT cry VFORM pres) • F1 and F2 do not unify: F1 = (CAT v AGR 3s) F2 = (CAT v AGR 3p)

  4. Unification-based Grammars • Grammar rules can be completely specified using unification • Example: X0 --> X1 X2 CAT X0 = S CAT X1 = NP CAT X2 = VP AGR0 =AGR1 = AGR2 VFORM0 = VFORM2 • If a feature (such as CAT) is always specified, it can be associated with the non-terminal of a CFG rule • Examples: S  NP VP AGR0= AGR1 = AGR2 VFORM0 = VFORM2 NP  ART N AGR0 = AGR1 = AGR2

  5. Sample Grammar (Abbreviated Form)

  6. Unification Grammars

  7. Feature Structures as DAGs • Feature Structures are commonly represented as DAGs • Unification between FSs can be described as an operation that constructs a new DAG representing the unified structure • the Algorithm:

  8. Representing FSs as DAGs • Feature Structures: N1: (CAT N ROOT fish AGR {3s 3p}) N2: (CAT N AGR 3s) are represented as:

  9. Graph Unification Algorithm • To unify a DAG rooted at node Ni with a DAG rooted at node Nj: • 1. If Ni equals Nj, then return Ni and succeed. • 2. If both Ni and Nj are sink nodes, then if their labels have a non-null intersection, return a new node with the intersection as its label. Otherwise, the DAGs do not unify. • 3. If Ni and Nj are not sinks, then create a new node N. For each arc labeled F leaving Ni to node NFi, • 3a. If there is an arc labeled F leaving Nj to node NFj, then recursively unify NFi and NFj. Build an arc labeled F from N to the result of the recursive call. • 3b. If there is no arc labeled F from Nj, build an arc labeled F from N to NFi. • 3c. For each arc labeled F from Nj to node NFj, where there is no F arc leaving Ni, create a new arc labeled F from N to NFj.

  10. Building new constituents using feature equations

  11. N1: (CAT N ROOT fish AGR {3s 3p}) ART1: (CAT ART ROOT the AGR {3s 3p}) NP  ART N CAT0 = NP CAT1 = ART CAT2 = N AGR = AGR1 = AGR2

  12. DAG Unification

  13. DAG Representation of is happy

  14. The fish is happy

  15. Predicative Phrases • VP  (V ROOT be) (NP PRED +) He is a student • VP  (V ROOT be) (PP PRED +) He is in the house • VP  (V ROOT be) (ADJP PRED +) He is happy • X0  X1 X2 CAT0 =VP CAT1 = V CAT2 = {NP PP ADJP} ROOT1 = be PRED2 = +

  16. More Information in Lexicon put : (CAT V SUBCAT (FIRST (CAT NP) SECOND (CAT PP LOC +))) VP  V X2 X3 2 = FIRSTSUBCAT1 3=SECONDSUBCAT1 want : (CAT V SUBCAT (FIRST (CAT NP) SECOND (CAT VP VFORM inf)))

More Related