CS4025: Semantics

1
CS4025 Semantics

• Representing meaning
• Semantic interpretation
• Word meaning
• For more information JM, chap 14, 16 in 1st ed
  17, 19 in 2nd

2
NL Understanding

• Understanding written text
• Which books are bestsellers
• Who wrote them
• For now, focus on AI approach
• explicit models of grammar, meaning, etc

3
Stages

• Morphology analyse word inflection
• Syntax determine grammatical structure
• Semantics convert to a form that is meaningful
  to a computer
• eg, SQL query
• Pragmatics influence of context
• eg, what them refers to

4
Example

• Original Who wrote them
• morph who write/past them
• Grammar verbwrite, subjectwho, objectthem
• semantics Select title, firstname, lastname from
  X
• pragmatics
• Select title, firstname, lastname from books
• Where salesthisyear gt10000

5
Definition

• Semantic interpretation rewrites a parse tree
  into a meaning representation
• Logic, SQL, knowledge base
• Poorly understood compared to syntax
• apps that need complex semantics, like database
  front ends or high-quality MT, have had limited
  success in the past

6
Meaning

• How can we represent the meaning of an English
  sentence?
• Programming languages meaning is the
  equivalent machine code
• a b c
• means load a
• add b
• store c
• We could represent meaning as programs in some
  language, in which case NLU would be a kind of
  compilation

7
Meaning Representation in NL

• Many possibilities
• executable programs
• logical formulas
• AI knowledge representation
• nothing
• No consensus on what is best - basic problem in
  philosophy and psychology

8
Criteria for an ideal MRL

• Unambiguous
• Able to express all necessary shades of meaning
  for the application domain
• Verifiability system can tell whether a
  statement is true according to a knowledge base
• Canonical different sentences with the same
  meaning are mapped to the same representation
• Support of inference

9
Example John passed CS1001

• Different representations
• Program C (or SQL) code to add an appropriate
  entry to a student database
• Logic pass(John, CS1001)
• AI Semantic Net

10
Program as representation

• Translate English into SQL (C, ...)
• MS English Query / AccessELF
• List the bestsellers translated into Select
  titles from books where salesgt10000
• Usually need a different translator for each
  application
• Good authoring environments for semantic rules
  are essential

11
Logic as a Representation

• Translate into (first-order) logic
• John is a man man(John)
• John eats spinach eat(John,spinach)
• John sold all of his stocks
• (?X)(stock(X) own(John,X)) ? sell(John, X))
• John sold Peter all of his stocks
• (?X)(stock(X) own(John,X)) ? sell(John,X,Peter))
  

12
Logic as Representation (2)

• Good points
• Can represent any meaning (if you are inventive
  enough about predicates etc.)
• Good support for compositionality, arbitrarily
  complex statements
• Good support for quantifiers (all, some,...)
• Bad points
• Doesnt seem to really match the way people
  think.
• does ? really mean some?

13
Case Frames as a Representation

• Form of (AI) semantic network
• Assume verbs (and other words) are objects with
  relations
• AGENT - the person/thing acting
• THEME - the person/thing acted upon
• BENEFICARY - of action
• AT-LOC - where action happened

14
Example

• John gave Peter the ball
• John gave the ball to Peter
• The ball was given to Peter by John
• are all interpreted as
• GIVE
• agent John
• theme ball
• beneficiary Peter

15
Meaning Primitives

• Meaning primitives are a fixed set of concepts/
  roles etc. in terms of which any meaning can be
  expressed
• Makes reasoning, e.g. about whether two meanings
  are the same, simpler.
• Example PURCHASE act
• John bought the book from Sam
• Sam sold the book to John
• Difficult to define small set of primitives
• Conceptual Dependency was one serious attempt

16
Conceptual Dependency

• Primitives
• ATRANS - abstract transfer
• PTRANS - physical transfer
• MTRANS - mental transfer
• PROPEL - apply force to an object
• INGEST - eat, drink, etc
• CON - conceptualise
• etc

17
Example "John bought a book from Mary."

• (BI-CAUSE
• (SOURCE (ATRANS (ACTOR MARY)
• (OBJECT BOOK)
• (FROM MARY)
• (TO JOHN)
• (TIME
  PAST)))
• (TARGET (ATRANS (ACTOR JOHN)
• (OBJECT
  MONEY)
• (FROM JOHN)
• (TO MARY)
  
• (TIME
  PAST))))

18
Example "Bob threw the ball to Bill."

• (PTRANS (ACTOR BOB)
• (OBJECT BALL)
• (FROM BOB)
• (TO BILL)
• (TIME PAST)
• (INSTRUMENT (PROPEL (ACTOR BOB)
• (OBJECT
  BALL)
• (FROM BOB)
• (TO
  BILL)
• (TIME
  PAST)))

19
Knowledge Bases

• Represent meaning using objects in a large AI
  knowledge base
• CYC project - 15-year project to build a
  knowledge base which holds the kind of general
  world knowledge that people have
• Use Cyc primitives and KR language to represent
  meaning?

20
MRLs and Logic

• Most existing meaning representation languages
• (frames, semantic nets, case frames etc). can be
  viewed as subsets of First Order Logic (where the
  expressive power is restricted or the set of
  predicates etc. is partially determined)
• Main deficiencies of first order logic
  inability to express default inferences and
  inferences based on partial information

21
Choosing an MRL What is the Task?

• Why are we processing this sentence? This could
  influence the kind of meaning representation
  chosen
• database interface - perhaps use SQL rep?
• AI system which reasons about Johns problems -
  perhaps use logic or AI KR?
• Information retrieval, speech dictation, grammar
  checking - dont build any meaning representation?

22
Semantic Interpretation

• Rewriting the parse tree into the target
  representation
• May be based on rewrite rules that insert a
  semantic structure X if the parse tree contains
  syntactic structure Y
• For generality/coverage, needs to be
  compositional, that is the meaning of the whole
  is some fixed function of the meanings of the
  parts
• More on this in the next lecture

23
Ex List the books

• S imperative
• V List
• NP X
• mapped into
• Select X.ltnamegt from X
• There are also cheaper/simpler approaches to
  semantic interpretation in use

24
Semantic interpretation Semantic grammar

• An attempt to reduce the distance between
  syntactic and semantic representations
• Grammar is defined in terms of semantic
  categories
• TIMEQ-gt When does FLIGHT-NP FLIGHT-VP
• FLIGHT-NP -gt Flight NUMBER
• FLIGHT-NP -gt Flight to CITY
• FLIGHT-NP -gt TIME flight to CITY
• FLIGHT-VP -gt depart
• FLIGHT-VP -gt leave

25
Semantic Interpretation Template spotting

• Look for patterns (either in text or parse tree)
  which identify meaning fragments
• Example How much is a ticket to London?
• How much specifies cost query
• a ticket specifies a single one-way ticket
• to London specifies destination
• Must be in limited domain
• Patterns looked for can be informed by knowledge
  about how words relate to underlying concepts and
  what syntactic properties words have.

26
Doctor-on-Board Problem

• Simple rewriting may not be sufficient. Example
• Is there a doctor within 200 miles of the
  Enterprise
• Database doesnt have Doctor entities, instead it
  has DoctorOnBoard attr for ships
• Need to rephrase this as
• Is there a ship within 200 miles of the
  Enterprise which has a doctor on board?
• Restructure query from humans data model to
  databases data model
• Distance between syntactic and semantic structure
  significant in this example

27
Lexical (Word) Meaning

• Logic (classical) model
• bachelor(X) male(X) adult(X) married(X)
• But the pope? Divorcee? Muslim with 3 wives?
• Father(X) male(X)parent(X)
• Man who adopts a child?
• Sperm-bank donor?
• Unmarried partner to woman raising a child?
• Unmarried (gay) partner to man raising a child?
• Prototype/exemplar models may be better when
  words dont have formal definitions

28
Word meaning for time-series data

• Weather reports
• Saturday will be yet another generally dull day
  with early morning mist or fog and mainly cloudy
  skies being prevalent. There will be the odd
  bright spell here and there, but it will feel
  rather damp with patches of mainly light rain to
  be found across many parts, especially the west
  and south.
• Ongoing research project in CS Dept

29
Current Applications and Tasks

• Meaning Representation Parsing SemEval-2016 Task
  8 http//alt.qcri.org/semeval2016/task8/
• CC/Boxer http//gmb.let.rug.nl/
• Attempto Controlled English http//attempto.ifi.u
  zh.ch/site/
• Would also want to do such semantic
  representations for Semantic Web applications.

30
Conclusion

• Converting sentences to a meaning
  representation is hard
• No agreement on best meaning-rep
• Word meaning is hard to pin down
• Limited success in small domains, but we cant
  semantically interpret general text
• but we can parse general text