31
Computational language: week 9 Finish finite state machines FSA’s for modelling word structure Declarative language models knowledge representation and inheritance non-monotonic or ‘default’ inheritance and lexical knowledge representation multiple inheritance DATR: a lexical knowledge representation language

Computational language: week 9 Finish finite state machines FSA’s for modelling word structure Declarative language models knowledge representation and

  • View
    214

  • Download
    0

Embed Size (px)

Citation preview

Computational language: week 9 Finish finite state machines

FSA’s for modelling word structure Declarative language models

knowledge representation and inheritance non-monotonic or ‘default’ inheritance and

lexical knowledge representation multiple inheritance DATR: a lexical knowledge representation

language

Finite State Automata Deterministic FSAs

An FSA whose recognition behaviour is fully determined by the state it is in and the input symbol it is looking at

Finite State Automata Deterministic FSAs

An FSA whose recognition behaviour is fully determined by the state it is in and the input symbol it is looking at

Non-deterministic FSAs An FSA with decision points

Finite State Automata Deterministic FSAs Non-deterministic FSAs

An FSA with decision points Self-loop may be in a particular state Arcs may have ε transitions

Finite State Automata Formal language Set of strings Finite symbol set, alphabet

Finite State Automata Formal language Set of strings Finite symbol set, alphabet L(m) = {baa!, ba!, baaa!,…}“formal language characterised by m”

m = model L = formal language

Finite State Automata Formal language Set of strings Finite symbol set, alphabet L(m) = {baa!, ba!, baaa!,…} A formal language models a

fragment of a natural language

Finite state transducers and word structure FSA for word structure

Handles the morphotactics Arc labels drawn from sets of lexical

classes Lexical entries and affixes in the

lexicon Achieve morphological recognition

Finite state transducers and word structure Finite State Transducers for

morphological parsing distinction between surface level and lexical level

lexical level includes basic stem and morphosyntactic features

cat +N+PL -> cats

Finite state transducers and word structure Finite State Transducers for

morphological parsing distinction between surface level and lexical level

lexical level includes basic stem and morphosyntactic features

cat +N+PL -> cats FST: maps one of set of symbols onto

another

Finite state transducers and word structure Finite State Transducers for

morphological parsing distinction between surface level and lexical

level lexical level includes basic stem and

morphosyntactic featurescat +N+PL -> cats FST: maps one of set of symbols onto

another c:c a:a t:t +N:εPL:s

Finite state transducers and word structure Finite State Transducers for

morphological parsing Boundary markers:

^ morpheme boundary # word boundary

Finite state transducers and word structure Finite State Transducers for

morphological parsing Boundary markers:

^ morpheme boundary # word boundary

c:c a:a t:t +N:εPL:^s#

Finite state transducers and word structure Finite State Transducers for

morphological parsing Boundary markers:

^ morpheme boundary # word boundary

c:c a:a t:t +N:εPL:^s# g:g o:e o:e s:s e:e +N:εPL:#

Knowledge representation Linguistic knowledge Linguistic facts Relations between linguistic facts Computable

Knowledge representation Linguistic knowledge Linguistic facts

lexicon Relations between linguistic facts Computable

Knowledge representation Linguistic knowledge Linguistic facts

lexicon Relations between linguistic facts

inference default inference

Computable

Knowledge representation Linguistic knowledge Linguistic facts

lexicon Relations between linguistic facts

inference default inference

Computable DATR Prolog

Knowledge representation and the lexicon

Inheritance hierarchy

Knowledge representation and the lexicon

Inheritance hierarchy A graphical representation of how

facts or properties are shared amongst entities

Knowledge representation and the lexicon

Inheritance hierarchy BIRD:

wings = yes flies = yes eggs = yes feathers = yes

Seabird: food = fish Landbird:

food = insects

Knowledge representation and the lexicon

Inheritance hierarchy BIRD:

wings = yes flies = yes eggs = yes feathers = yes

Seabird: food = fish Landbird:

food = insects

Seagull Goldfinch

Knowledge representation and the lexicon Inheritance hierarchy

BIRD:

wings = yes flies = yes eggs = yes feathers = yes

Seabird: food = fish Landbird:

food = insects

Seagull Goldfinch

Penguin?

Knowledge representation and the lexicon

Default inheritance hierarchy BIRD:

wings = yes flies = yes eggs = yes feathers = yes

Seabird: food = fish Landbird:

food = insects

Seagull Goldfinch

Penguin: flies = no

Knowledge representation and the lexicon

Default inheritance

NOUN: singular = plural = s

Count_noun Noncount_noun:

plural = undefined

Book Sand

Cat

Knowledge representation and the lexicon

Default inheritance

NOUN: singular = plural = s

Count_noun Noncount_noun:

plural = undefined

Book Sand

Cat

Ox?

Knowledge representation and the lexicon

Default inheritance

NOUN: singular = plural = s

Count_noun Noncount_noun:

plural = undefined

Book Sand

Cat

Ox: plural = en

Knowledge representation and the lexicon

Default inheritance

NOUN: singular = plural = s

Count_noun Noncount_noun:

plural = undefined

Book Sand

Cat

Ox: plural = en

Goose: plural = geese

Knowledge representation and the lexicon

DATR, lexical knowledge representation language

Inference and default inference Compiled in Prolog Developed by Evans and Gazdar Network Morphology

Knowledge representation and the lexicon

DATR demonstration

Lexical knowledge representation

Summary

Inheritance Inference rules Formalism Computable