Natural Language Processing Lecture 6 Augmented Transition Networks Reading: James Allen NLU (Chapter 4)

Augmented Transition Networks An augmented transition network (ATN) is a recursive transition network that can perform tests and take actions during arc transitions. An ATN uses a set of registers to store information. A set of actions is defined for each arc, and the actions can look at and modify the registers. An arc may have a test associated with it. The arc is traversed (and its action is taken) only if the test succeeds. When a lexical arc is traversed, it is put in a special variable (*) that keeps track of the current word.

An ATN for NP

An ATN for S

1 the 2 dog 3 saw 4 Jack 5

An ATN Grammar for Simple Declarative Statements

The NP Network

The PP Network

Definite Clause Grammars

S(np(art(the), n(dog)), vp(v(cried)))

Definite Clause Grammars An convenient abbreviated format for writing grammar rules for logic-based parsing systems Rules can then automatically be compiled into a full database of PROLOG clauses Omit technical arguments that can automatically be inserted by the system: – word positions and constituent spans – explicit predicates for the lexicon entries Resulting grammar format looks something like: S(s(Np,Vp),FS) -> NP(Np,FS), VP(Vp,FS) NP(np(Art,Noun),FS) -> ART(Art,FS), N(Noun,FS) VP(vp(Verb),FS) -> V(Verb,FS)

A declarative clause grammar

Register Presetting Register Presetting  Parameter Passing vp:inf Mary wants to have a party. _np_vp:inf Mary wants John to have a party. Register VFORM of VP Graph should be preset to inf