CSC 594 Topics in AI –
Natural Language Processing
Spring 2016/17
14. Semantic Representation
(Some slides adapted from Jurafsky & Martin)
1
Meaning Representations
• We’re going to take the same basic approach to
meaning that we took to syntax and morphology
• We’re going to create representations of linguistic
inputs that capture the meanings of those inputs.
• But unlike parse trees, these representations
aren’t primarily descriptions of the structure of
the inputs…
Speech and Language Processing - Jurafsky and Martin 2
Semantic Processing
• Ok, so what does that mean?
• Representations that
– Permit us to reason about their truth (i.e., their relationship to
some world)
– Permit us to answer questions based on their content
– Permit us to perform inference (answer questions and determine
the truth of things we don’t already know to be true)
Speech and Language Processing - Jurafsky and Martin 3
Semantic Processing
• Several ways to attack this problem
– Limited, shallow, practical approaches that have some hope of
actually being useful
• Information extraction
– Principled, theoretically motivated approach…
• Computational/Compositional Semantics
– Chapters 17 and 18
– Something midway that can plausibly serve both purposes
• Semantic role labeling
Speech and Language Processing - Jurafsky and Martin 4
Semantic Analysis
• Compositional Analysis
– Create a logical representation that accounts for all the entities,
roles and relations present in a sentence.
Speech and Language Processing - Jurafsky and Martin 5
Representational Schemes
• We’re going to make use of First Order Logic (FOL) as
our representational framework
– Not because we think it’s ideal
– Many of the alternatives turn out to be either too limiting or
– They turn out to be notational variants
Speech and Language Processing - Jurafsky and Martin 6
FOL
• Allows for…
– The analysis of truth conditions
• Allows us to answer yes/no questions
– Supports the use of variables
• Allows us to answer questions through the use of variable binding
– Supports inference
• Allows us to answer questions that go beyond what we know
explicitly
Speech and Language Processing - Jurafsky and Martin 7
Meaning Structure of Language
• Natural languages convey meaning through the use of
– Predicate-argument structures
– Variables
– Quantifiers
– Compositional semantics
Speech and Language Processing - Jurafsky and Martin 8
Predicate-Argument Structure
• Events, actions and relationships can be captured with
representations that consist of predicates and arguments
to those predicates.
• Languages display a division of labor where some words
and constituents (typically) function as predicates and
some as arguments.
Speech and Language Processing - Jurafsky and Martin 9
Predicate-Argument Structure
• Predicates
– Primarily Verbs, VPs, Sentences
– Sometimes Nouns and NPs
• Arguments
– Primarily Nouns, Nominals, NPs, PPs
Speech and Language Processing - Jurafsky and Martin 10
Example
• Mary gave a list to John.
• Giving(Mary, John, List)
• More precisely
– Gave conveys a three-argument predicate
– The first argument is the subject
– The second is the recipient, which is conveyed by the NP inside
the PP
– The third argument is the thing given, conveyed by the direct
object
Speech and Language Processing - Jurafsky and Martin 11
Note
• Giving(Mary, John, List) is pretty much the same as
– Subj(Giving, Mary), Obj(Giving, John), IndObj(Giving, List)
– Which should look an awful lot like.... what?
Speech and Language Processing - Jurafsky and Martin 12
Better
• Turns out this representation isn’t quite as useful as
it could be.
• Better would be
Speech and Language Processing - Jurafsky and Martin 13
Predicates
• The notion of a predicate just got more complicated…
• In this example, think of the verb/VP providing a
template like the following
• The semantics of the NPs and the PPs in the sentence
plug into the slots provided in the template
Speech and Language Processing - Jurafsky and Martin 14
Two Issues
• How can we create this • What makes that
kind of representation in representation a
a principled and efficient “meaning”
way representation, as
opposed say to a parse
tree?
Speech and Language Processing - Jurafsky and Martin 15
Semantic Analysis
• Semantic analysis is the process of taking in some
linguistic input and assigning a meaning representation
to it.
– There a lot of different ways to do this that make more or less
(or no) use of syntax
– We’re going to start with the idea that syntax does matter
• The compositional rule-to-rule approach
Speech and Language Processing - Jurafsky and Martin 16
Compositional Analysis
• Principle of Compositionality
– The meaning of a whole is derived from the meanings of the
parts
• What parts?
– The constituents of the syntactic parse of the input
• What could it mean for a part to have a meaning?
Speech and Language Processing - Jurafsky and Martin 17
Example
• Franco likes Frasca.
Speech and Language Processing - Jurafsky and Martin 18
Compositional Analysis
Speech and Language Processing - Jurafsky and Martin 19
Augmented Rules
• We’ll accomplish this by attaching semantic formation
rules to our syntactic CFG rules
• Abstractly
• This should be read as the semantics we attach to A
can be computed from some function applied to the
semantics of A’s parts.
Speech and Language Processing - Jurafsky and Martin 20
Example
• Easy parts… • Attachments
– NP -> PropNoun {PropNoun.sem}
– PropNoun -> Frasca {Frasca}
– PropNoun -> Franco {Franco}
Speech and Language Processing - Jurafsky and Martin 21
Lambda Forms
• A simple addition to
FOL
– Take a FOL sentence
with variables in it that
are to be bound.
– Allow those variables
to be bound by
treating the lambda
form as a function
with formal arguments
Speech and Language Processing - Jurafsky and Martin 22
Compositional Semantics by
Lambda Application
Speech and Language Processing - Jurafsky and Martin 23
Lambda Applications and
Reductions
Speech and Language Processing - Jurafsky and Martin 24
Lambda Applications and
Reductions
VP VP Verb NP {Verb.sem(NP.Sem)
Frasca
Speech and Language Processing - Jurafsky and Martin 25
Complex NPs
• Things get quite a bit more complicated when we start
looking at more complicated NPs
– Such as...
• A menu
• Every restaurant
• Not every waiter
• Most restaurants
• All the morning non-stop flights to Houston
Speech and Language Processing - Jurafsky and Martin 26
Quantifiers
• Contrast...
– Frasca closed
• With
– Every restaurant closed
Speech and Language Processing - Jurafsky and Martin 27
