NLP Module - 2

Outline
 Cohesion and Coherence
 Ambiguity
 Natural Language Generation
Cohesion & Coherence
Cohesion
 Definition 1:
 isthe grammatical and lexical relationship within a
text or sentence and can be defined as the links
that hold a text together and give it meaning.
 Cohesion is when the link between sentences,
words and phrases are visible, or easily
understandable.
 Ref
 http://en.wikipedia.org/wiki/Cohesion_(linguistics)
Cohesion
 Definition 2:
 Cohesion is a semantic relation between an
element in the text and some other element
that is crucial to the interpretation of it.
 Ref
 Halliday et al., 1976
 Example:
 Ahmad belongs to Peshawar. He is in M.Sc
final year
Coherence
 Definition 1:
 A quality of sentences, paragraphs, and essays
when all parts are clearly connected.
 Coherence is when the theme or the main idea
of the essay or writing piece is understandable.
A text has coherence if its constituent
sentences follow on one from the other in an
orderly fashion so that the reader can make
sense of the entire text.
 Ref(http://grammar.about.com/od/c/g/coherenceterm.htm: retrieved: 30
Oct, 2010)
Coherence
 Example 1:
 There once was a farmer in a small village.
He worked hard day and night in his fields
to fed his wife and children.
 Example 2:
 Aliwas a studious student and got 900
marks. His percentage is 92.1%.
Coherence
 Definition 2:
 Coherence is a semantic property of discourse
formed through the interpretation of each
individual sentence relative to the interpretation
of other sentences, with "interpretation" implying
interaction between the text and the reader.
 Ref(Teun A. van Dijk, pp. 93)
 http://www.criticism.com/da/coherence.php
Coherence
 Definition 3:
 Coherence in linguistics is what makes a text semantically
meaningful.
 It is especially dealt with in text linguistics.
 http://en.wikipedia.org/wiki/Coherence_(linguistics), retrieval date: 27 Oct, 2010

 Definition 4:
 When sentences, ideas, and details fit together clearly,
readers can follow along easily, and the writing is coherent.
http://home.ku.edu.tr/~doregan/Writing/Cohesion.html, Retrieved 30 Oct, 2010
Why coherence?
 The text-based features which provide cohesion in a
text do not necessarily help achieve coherence, that
is, they do not always contribute to the
meaningfulness of a text, be it written or spoken.
 It has been stated that a text coheres only if the world
around is also coherent.
Cohesive devices
 Definition:
 The links within the text that hold it
together are called cohesive devices.
Categories of cohesive devices
 A cohesive text is created in many different ways. In
Cohesion in English, M.A.K. Halliday and Ruqaiya
Hasan (1976) identify five general categories of
cohesive devices that create coherence in texts:
 Reference,
 Ellipsis,
 Substitution,
 Conjunction,
 Lexical cohesion.

 Refhttp://www.criticism.com/da/coherence.php
References or referring expressions

 A natural language expression used to perform reference is

called a referring expression and the entity that is referred to is
called the referent.
 Referring expressions are words or phrases, the semantic interpretation of
which is a discourse entity (also called referent)
 Example:
 A pretty woman entered the restaurant. She sat at the table
next to mine and only then I recognized her. This was Amy
Garcia, my next door neighbor from 10 years ago. The woman
has totally changed! Amy was at the time shy…
References
A pretty woman entered the restaurant. She sat at the
table next to mine and only then I recognized her.
This was Amy Garcia, my next door neighbor from
10 years ago. The woman has totally changed! Amy
was at the time shy…

Referent

Anaphoric Referring Cataporic Referring

Expression Expression
Types of references

We can summaries reference with a diagram to

make it easier to grasp:
Exophoric reference
 Exophoric reference, depends on the context outside the text
for its meaning.
 In linguistics, Exophora is reference to something extra-
linguistic.
 For example
 "What is this?",
 here "this" is exophoric rather than endophoric, because it
refers to something extra-linguistic, i.e. there is not enough
information in the utterance itself to determine what "this"
refers to, but we must instead observe the non-linguistic
context of the utterance (e.g. the speaker might be holding an
unknown object in their hand as they ask that question.)
Endophoric reference
 The pronouns refer to items within the same text; it is
endophoric reference.
 Endophora is a linguistic reference to something intra-
linguistic.
 For Example:
 "I saw Sally yesterday. She was lying on the beach".
 Here, "she" is intra-linguistic, and hence endophoric, because it refers to
something (Sally, in this case) already mentioned in the text.

 (From Wikipedia, the free encyclopedia)

Endophora (Another Definition)
 Words or phrases like pronouns are endophora
when they point backwards or forwards to
something in the text:
 For example:
 As [he]1 was late, [Harry] 1 wanted to phone [his] 1

[boss] 2 and tell [her] 2 what had happened.

 (From Wikipedia, the free encyclopedia)

Types of Endophora
 1) Cataphora:
 The type of endophora in which the referring expression occur
before the referent are termed as cataphora.

 OR The type of endophora in which the pronouns link forward

to a referent (nouns) in the text that follows.
 For example:
 When [she] 1 saw the snake, [Harry] 1 cried.
 The elevator opened for [him] 1 on the 14th flour, and [Ali] 1
stepped out quickly.
 2) Anaphora:
 The type of endophora in which the referent occurs before
the referring expression are termed as anaphora.

 OR The type of endophora in which the pronouns link backward

to a referent (nouns) in the text.
 For example:
Anaphora(Another Definition)
 Anaphora is a phenomenon in which certain textual
elements refer to earlier text elements (called
correlates) and share the meaning of the correlates.

 For Example:
 1)John helped Mary.
Correlate/
 2) He was kind.
Referent/

Referring Element/ Antecedent

Anaphor/
Anaphoric Device (AD)
Anaphoric and Cataphoric Devices
 The referring elements (pronouns) in anaphoric text that refer to their
corresponding referent ( nouns) backward are called anaphoric
devices. Also, called anaphor.
 For example:
 Bell is a powerful player but unfortunately he will not take part in the
trophy due to injury.
Anaphoric Device
 The referring elements (pronouns) that refer to their corresponding referent
(nouns) forward in cataphoric text are called cataphoric devices.
Also, called cataphor.
 For example:
 As her father went abroad, Nighat took control of the organization by
herself.
Cataphoric Device
Antecedent
 The referent in the anaphoric/cataphoric text to which the
anaphoric/cataphoric devices refer are called antecedents.
 Also, called correlates .
 For example:
 Bell is a powerful player but unfortunately he will not
take part in the trophy due to injury.

Antecedent
 As her father went abroad, Nighat took control of the
organization by herself.
Types of Anaphora (On the basis of position of
anaphor and its antecedent)
 Intra-sentential/Sentence internal anaphora:
 The anaphora in which the AD and its antecedent both
occurs in the same sentence is called sentence internal.
 Reflexive pronouns
 (himself, herself, itself, themselves) are typical examples of intra-
sentential anaphora.
 Possessive pronouns
 (his, her, hers, its, their, theirs) can often be used as intra-sentential
anaphors too, and often be in the same clause as the anaphor.
 For example:
 [John] 1 took [his] 1 [hat] 2 off and hung [it] 2 on a peg.
Types of Anaphora (Cont..)
 Inter-sentential/Sentence external anaphora:
 The anaphora in which the AD and its antecedent
doesn’t occur in the same sentence is called
sentence external or inter-sentential anaphora.
 For example:
 [Jehansher] 1 Khan was senior player of Sqash. [He] 1 has
won several trophies.
 [John] 1 took his hat off and hung it on a peg. [He] 1 was very
tied therefore went to slept..
Types of Anaphora
(On the basis of grammatical category of the anaphor and antecedent)

 Noun phrased anaphora

 Personal anaphora
 Non-personal
 Verb phrased anaphora
 Pronominal anaphora
 Definite noun phrase
 Quantifier/Ordinal anaphora
 Zero anaphora
 Identity-of-reference anaphora
 Identity-of-sense anaphora
Anaphora and ambiguity
 Many anaphors are ambiguous. Like:
 A)
 Jane told Marry she was in love (ambiguous)
 Jane informed Marry she was in love. (Here Jane

is in love)
 B)
 Jane told marry she was in danger (ambiguous)
 Jane warned Marry she was in danger.
Anaphora resolution
 Anaphora Resolution == the problem of resolving what a
pronoun, or a noun phrase refers to.
 Consider the following Discourse:

1) John helped Mary.

2) He was kind.

 After anaphora resolution:

1) John helped Mary.

2) John was kind.
Ellipsis
 Definition:
 The omission of a portion, of a phrase or a
sentence is called Ellipsis (Rav, L., F.).
 Example:

ᶲ
He is rich, but his brother is not .

Bob ᶲ and Tom ate cheese.
Other examples
 George bought a huge box of chocolates but few Ǿ were
left by the end of the day.
Chocolates

 I have never been to Karachi but my father has Ǿ , and he

says it was wonderful.

Been to Karachi
Origin of the word ellipsis
 Derived from Greek, the word 'ellipsis' means
“the omission of words that could be
understood from the context”.
 Ellipsis is the non-expression of one or more
sentence elements whose meaning can be
reconstructed either from the context or from a
person’s knowledge of the world.
e-clause & a-clause
 e-clause:
 The clause from which the material is missing is often
referred to as the elliptical clause (e-clause).
 Example:

He is rich, but his brother is not . ᶲ
 a-clause:
 The clause from which the interpretation of the missing
material is derived is referred to as antecedent clause (a-
clause).
 Example:

ᶲ
He is rich, but his brother is not .
Types of ellipsis
 Some of the several types of ellipses are:
 Noun Phrase Ellipses
 Verb Phrase Ellipses
 Gapping
 Stripping
 Sluicing
 Ellipses in WH-Constructions
 Ellipses in Q-Constructions
Ellipsis resolution
 Ellipses resolution is an important area in the
research community.
 All natural languages have the occurrences of
ellipses in their text as well as speech,
although different from each other.
 For their resolution different approaches are
followed by linguists as described by Shalom
Lappin (Lappin and Lease).
Examples of ellipsis resolution
 He is rich, but his brother is not ᶲ.
 (= He is rich but his brother is not rich).
 Bob ᶲ and Tom ate cheese.
(= Bob (ate cheese) and Tom ate cheese).
 John realizes that he is a fool, but Bill does not ᶲ,
even though his wife does ᶲ.
 (= John realizes that John is a fool, but Bill does not
realize that Bill is a fool, even though Bill's wife
realizes that Bill is a fool.)
Substitution
 Substitution
 Substitution is very similar to ellipsis in the effect
it has on the text, and occurs when instead of
leaving a word or phrase out, as in ellipsis, it is
substituted for another, more general word.
 {example}
 "Which ice-cream would you like?“

 "I would like the pink one"

 "one" is used instead of repeating "ice-cream."
Conjunction
 Conjunction
 Conjunction creates cohesion by relating sentences
and paragraphs to each other by using words from
the class of conjunction, or numerals.
 This can be:
 temporal (after, before), He came home after/before me.
 causal (because), He was fired because he was chronically
late.
 coordinating (and), I had breakfast and left for work.
 adversative (but, still, yet, whereas, while, nevertheless),
The rope was thin but it was strong.
 additive (further) or
 discourse markers (now, well, after all).
Lexical cohesion
 Lexical Cohesion
 Lexical cohesion is basically created by repetition
(reiteration) of the same lexeme, or general (aka
shell) nouns, or other lexemes sharing the majority
of semantic features: The bus ... - the vehicle ... -
the chassis ....
 For example,
 'it', 'neither' and 'this'
 all refer to an idea previously mentioned.
 'First of all', 'then' and 'after that'
 help to sequence a text.
 'However', 'in addition' and 'for instance'
 link ideas and arguments in a text.
Summary of cohesion
 Cohesion is the glue that holds a piece of writing
together.
 In other words, if a paper is cohesive, it sticks together
from sentence to sentence and from paragraph to
paragraph.
 Cohesive devices certainly include transitional words and
phrases, such as:
 therefore, furthermore, or for instance, that clarify for
readers the relationships among ideas in a piece of writing.
However, transitions aren't enough to make writing cohesive.
 Repetition of key words and use of reference words are also
needed for cohesion.
Cohesion vs coherence
 Difference 1:
 Coherent:
 The semantic relationships which link the meanings of utterances in a discourse.
 Cohesion:
 The grammatical and lexical relationships between the different elements of the
text in a discourse.
 Difference 2:
 Coherence:
 Coherence is a super set which means that if a text is coherent then it will also be
cohesive but the converse is not true
 Cohesion:
 It is a sub set of coherence which means that if a text is cohesive then it may or
may not be coherent.
 Difference 3:
 Coherence:
 The relationship among the parts of the text and sentences is strong.
 Cohesion:
 The relationship among the parts of the text and sentences is casual.
Bibliography
 Daniel Jurafsky, Speech and Language Processing: An Introduction to Natural
Language Processing, Computational Linguistics and Speech Recognition, Pearson
Education,Inc, 1999, Chapter No. 18 .

 Halliday, M.A.K; and Ruqayia Hasan (1976): Cohesion in English. London:

Longman.

 Rav, L., F. “The Understanding and Generation of Ellipsis in a Natural Language

System”, Berkeley AI Research Project Comp: Sc Division University of California
Berkeley California.

 M.A.Khan, “ Text Based Machine Translation System”, PhD Thesis, 1995.


Lappin, S., “A Sequenced Model of Anaphora and Ellipsis Resolution”, 2003 .
Ambiguity
Ambiguity
 Natural languages are inherently ambiguous.

 Humans generally manage to pick out the intended

meaning from a set of possible interpretations, but
hard for them as well in some situations.

 Due to their limited ‘knowledge’ and inability to get

their bearings in complex situations, understanding
natural language is a difficult task for computers.
Ambiguity: A Definition
 An input is termed to be ambiguous if there
are multiple alternative linguistic structures
that can be built for it.
 For example:
 “I made her duck”.
 This simple sentence could have five meanings.
Ambiguity Example

1. I cooked a duck for her.

1. I cooked a duck belonging to her.
2. I created the plaster duck she owns.
3. I caused her to quickly lower her head or
body.
4. I turned her into a duck (by magic).
Ambiguity (cont.)
These meanings are caused by a number of
ambiguities, i.e.
1. The words her and duck are syntactically
ambiguous.
2. The word make is semantically ambiguous
3. ‘I’ or ‘eye’ and ‘made’ or ‘maid’ (if the sentence is
a spoken one)
Types of Ambiguities
 Following are the various types of ambiguities:
 Lexicalor Semantic ambiguity
 Syntactic or Structural ambiguity
 Pragmatic ambiguity
 Referential ambiguity
Lexical Ambiguity (semantic ambiguity)
 An ambiguity caused when a word has more
than one meaning.
 For example:
 ‘The wrestlers entered the ring’.
 Ring can be a proposed area for wrestling.
 Ring can be a golden ring wrapped around a finger.
Lexical Ambiguity (cont.)

Following are some phenomenon of lexical

ambiguity:
1. Homonymy
2. Polysemy
3. Synonymy
4. Hyponymy
(Have already covered)
Syntactic Ambiguity
 Syntactic ambiguity occurs when a phrase or sentence
has more than one underlying structure.
 Syntactic ambiguity arises not from the range of
meanings of single words, but from the relationship
between the words and clauses of a sentence, and the
sentence structure implied thereby.
Syntactic Ambiguity (Cont..)
 Consider the following phrases and sentences:
 ‘Pakistani history teacher',
 'short men and women‘,
 'The girl hit the boy with a book‘,
 'Visiting relatives can be boring'.
 These ambiguities are said to be structural
because each such phrase can be represented in
two or more structurally different ways.
Syntactic Ambiguity (Cont..)
 The possible meanings for ‘Time flies like an
arrow’ are:
 1. Time goes by as quickly as an arrow.
 2. Time flies (N+PL) likes an arrow.
 3. Every "time fly" (N+SL) like an arrow.
 4. Someone or something named "Time" flies in a
way similar to an arrow. E.g.
 "Mr. Time flies an airplane fast and straight."
Pragmatic Ambiguity
 Occurs when speaker and listener don’t agree on
same principals of co-operative communication.
 For example:
 ‘I’ll meet you next Friday’ (which Friday)
 ‘Do you know the time?’
 ‘Can you pass the salt?’
etc.
Referential Ambiguity
 May be unclear what objects are referred to,
e.g.,
 1) He drove the car over the lawn mower, but it
wasn't hurt.
 2) Bob waved to Jim in the hallway between class.
He smiled.
Reading/References
 Daniel Jurafsky, Speech and Language Processing: An Introduction to Natural
Language Processing, Computational Linguistics and Speech Recognition, Pearson
Education,Inc, 2000 .

 M.A.Khan, “ Text Based Machine Translation System”, PhD Thesis, 1995.

Natural Language
Generation
Definition 1: NLG
 Natural Language Generation (NLG) is the process of
constructing natural language outputs from non-
linguistic inputs.
 Goal:
 The goal of this process can be viewed as the inverse of that
of natural language understanding (NLU)
 NLU Vs NLG:
 NLG maps from meaning to text, while NLU maps from text
to meaning.
(Juraffsky, Chapter 20)
Definition 2: NLG
 Natural Language Generation (NLG) is the
natural language processing task of generating
natural language from a machine
representation system such as a knowledge
base or a logical form.
(http://en.wikipedia.org/wiki/Natural_language_generation, Retrieved: 31 Oct,
2010)
Definition 3: NLG

Natural language generation is the process of

deliberately constructing a natural language text
in order to meet specified communicative goals.

[McDonald 1992]
What is NLG? Or Ingredient of NLG
 Goal:
 Computer software which produces understandable and
appropriate texts in English or other human languages
 Input:
 Some underlying non-linguistic representation of information
 Output:
 Documents, reports, explanations, help messages, and other
kinds of texts
 Knowledge sources required:
 Knowledge of target language and of the domain
Example System #1: FoG
 Function:
 Produces textual weather reports in English and French
 Input:
 Graphical/numerical weather depiction
 User:
 Environment Canada (Canadian Weather Service)
 Developer:
 CoGenTex
 Status:
 Fielded, in operational use since 1992
FoG: Input
FoG: Output
Example System #2: PlanDoc
 Function:
 Produces a report describing the simulation options that an engineer
has explored
 Input:
 A simulation log file
 User:
 Southwestern Bell Telephone Company (Texas)
 Developer:
 Bellcore and Columbia University
 Status:
 Fielded, in operational use since 1996
PlanDoc: Input
RUNID fiberall FIBER 6/19/93 act yes
FA 1301 2 1995
FA 1201 2 1995
FA 1401 2 1995
FA 1501 2 1995
ANF co 1103 2 1995 48
ANF 1201 1301 2 1995 24
ANF 1401 1501 2 1995 24
END. 856.0 670.2
PlanDoc: Output
This saved fiber refinement includes all DLC
changes in Run-ID ALLDLC. RUN-ID FIBERALL
demanded that PLAN activate fiber for CSAs 1201,
1301, 1401 and 1501 in 1995 Q2. It requested
the placement of a 48-fiber cable from the CO to
section 1103 and the placement of 24-fiber cables
from section 1201 to section 1301 and from
section 1401 to section 1501 in the second quarter
of 1995. For this refinement, the resulting 20 year
route PWE was $856.00K, a $64.11K savings over the
BASE plan and the resulting 5 year IFC was $670.20K,
a $60.55K savings over the BASE plan.
Example System #3: STOP
 Function:
 Produces a personalized smoking-cessation leaflet
 Input:
 Questionnaire about smoking attitudes, beliefs, history
 User:
 NHS (British Health Service)
 Developer:
 University of Aberdeen
 Status:
 Undergoing clinical evaluation to determine its effectiveness
STOP: Input
SMOKING QUESTIONNAIRE
Please answer by marking the most appropriate box for each question like this: 

Q1 Have you smoked a cigarette in the last week, even a puff?

YES  NO 
Please complete the following questions Please return the questionnaire unanswered in the
envelope provided. Thank you.

Please read the questions carefully. If you are not sure how to answer, just give the best answer you can.

Q2 Home situation:
Live  Live with  Live with  Live with 
alone husband/wife/partner other adults children

Q3 Number of children under 16 living at home ………………… boys ………1……. girls

Q4 Does anyone else in your household smoke? (If so, please mark all boxes which apply)
husband/wife/partner  other family member  others 

Q5 How long have you smoked for? …10… years

Tick here if you have smoked for less than a year 
STOP: Output
Dear Ms Cameron

Thank you for taking the trouble to return the

smoking questionnaire that we sent you. It appears
from your answers that although you're not planning
to stop smoking in the near future, you would like
to stop if it was easy. You think it would be
difficult to stop because smoking helps you cope
with stress, it is something to do when you are
bored, and smoking stops you putting on weight.
However, you have reasons to be confident of success
if you did try to stop, and there are ways of coping
with the difficulties.
Example System #4: TEMSIS
 Function:
 Summarizes pollutant information for environmental officials
 Input:
 Environmental data + a specific query
 User:
 Regional environmental agencies in France and Germany
 Developer:
 DFKI GmbH
 Status:
 Prototype developed; requirements for fielded system being
analyzed
TEMSIS: Input Query
((LANGUAGE FRENCH)
(GRENZWERTLAND GERMANY)
(BESTAETIGE-MS T)
(BESTAETIGE-SS T)
(MESSSTATION \"Voelklingen City\")
(DB-ID \"#2083\")
(SCHADSTOFF \"#19\")
(ART MAXIMUM)
(ZEIT ((JAHR 1998)
(MONAT 7)
(TAG 21))))

70
TEMSIS: Output Summary
 Le 21/7/1998 à la station de mesure de
Völklingen -City, la valeur moyenne maximale
d'une demi-heure (Halbstundenmittelwert) pour
l'ozone atteignait 104.0 µg/m³. Par conséquent,
selon le decret MIK (MIK-Verordnung), la valeur
limite autorisée de 120 µg/m³ n'a pas été
dépassée.
 Der höchste Halbstundenmittelwert für Ozon an
der Meßstation Völklingen -City erreichte am
21. 7. 1998 104.0 µg/m³, womit der gesetzlich
zulässige Grenzwert nach MIK-Verordnung von 120
µg/m³ nicht überschritten wurde.

71
Types of NLG Applications
 Automated document production
 weather forecasts, simulation reports, letters, ...
 Presentation of information to people in an
understandable fashion
 medical records, expert system reasoning, ...
 Teaching
 information for students in CAL systems
 Entertainment
 jokes (?), stories (??), poetry (???)
An Architecture for Generation

(Juraffsky, Chapter 20)

An Architecture for Generation (Cont..)
 Discourse Planner –
 This component starts with a communicative goal
and makes all the choices.
 It selects the content from the knowledge base
and then structures that content appropriately.
 The resulting discourse plan will specify all the
choices made for the entire communication,
potentially spanning multiple sentences and
including other annotations (including hypertext,
figures, etc.).
An Architecture for Generation (Cont..)
 Surface Realizer –
 This component receives the fully specified
discourse plan and generates individual
sentences as constrained by its lexical and
grammatical resources.
 These resources define the realizer's
potential range of output.
 If the plan specifies multiple-sentence
output, the surface realizer is called multiple
times.
Component Tasks in NLG
1. Content Determination
what information should be conveyed?
2. Discourse Planning
order & structure of message set
3. Sentence Aggregation
grouping messages into sentences
4. Lexicalization
words & phrases for concepts, relations
5. Referring Expression Generation
words & phrases for entities
6. Linguistic Realization
syntax, morphology, orthography
 Typical 3-Module/Pipelined Architecture
goal
1.Content Determination
Text
2. Discourse Planning
Planner
text plan
3. Sentence Aggregation
4. Lexicalization Sentence Q: How should these
5. Referring Expressions Planner be represented?

sentence plans
6. Syntax, Morphology, Linguistic
Orthography
Realizer
surface text
Text Plans
 Common representation : tree
 Leaf nodes = messages
 Internal nodes = message groupings
 Simple text plans: templates OK
 Complex text plans: require full representation
language
(e.g., TAMERLAN, DIOGENES)
Sentence Plans
 Simple: templates (select & fill)
 Complex: abstract representation
(SPL: Sentence Planning Language)
 Example SPL Expression
(S1/exist
:object (01/train
:cardinality 20
:relations ((R1/period
:value daily)
(R2/source
:value Aberdeen)
(R3/destination
:value Glasgow))))
There are 20 trains a day from Aberdeen to Glasgow
Content Determination
 Messages (raw content)
 User Model (influences content)
 Is Reasoning Required?
Find a train from Aberdeen to Leeds
(It requires two trains to get there)
 Deep Reasoning Systems
 represent the user’s goals as well as any
immediate query
 utilize plan recognition & reasoning
Discourse Planning
 Structure messages into a coherent text
 Example: start with a summary, then give
details
 Discourse relations, e.g.:
 elaboration:More specifically, X
 exemplification: For example, X
 contrast / exception: However, X
 Rhetorical Structure Theory (RST)
Sentence Aggregation
 No aggregation (1 sentence / message)
 Relative Clause
..which leaves at 10am
 Conjunction
..and the next train is the express
 Combinations
..and the next train is the express
which leaves at 10am
Lexicalization
 Choosing words to realize concepts or
relations
 Example:
(action/change
(measure outside_temperature)
(delta (quantity/deg_F -10)))

The temperature dropped 10 degrees

Lexical Selection Rules
(*A-INGEST
(AGENT *O-BOB)
(PATIENT *O-MILK)) => "drink"

(*A-INGEST
(AGENT *O-BOB)
(PATIENT *O-CHOCOLATE)) => "eat"
 Case Creation
• Additional structure is required to
realize the meaning of the
semantic representation
(*A-KICK
(AGENT *O-JOHN)
(PATIENT *O-BALL))

"John propelled the ball with his foot"

 Case Absorption
• Word chosen to realize a semantic
head also implies the meaning
conveyed by a semantic role
(*A-FILE-LEGAL-ACTION
(AGENT *O-BOB)
(PATIENT *O-SUIT)
(RECIPIENT *O-ACME))

"Bob sued Acme"

Referring Expression Generation
 Initial introduction
A man in the park looked up
 Pronouns
He saw a bird fly over
 Definite Descriptions
The man covered his head with a
newspaper
 Fixing Robot Text
• Start [the engine]i and run [the
engine]i until [the engine]i reaches
normal operating temperature
• Start []i and run [the engine]i until
[it]i reaches normal operating
temperature
• Second example introduces ellipsis
and anaphora
 Journalistic Style
“A dissident Spanish priest was charged here today
with attempting to murder the Pope. Juan Fernandez
Krohn, aged 32, was arrested after a man armed with
a bayonet approached the Pope while he was saying
prayers at Fatima on Wednesday night. According to
the police, Fernandez told the investigating magistrates today, he
trained for the past six months for the assault. If found guilty, the
Spaniard faces a prison sentence of 15-20 years.”
(Brown and Yule, 1983)
Reading/References
 Daniel Jurafsky, Speech and Language Processing: An Introduction to Natural Language
Processing, Computational Linguistics and Speech Recognition, Pearson Education,Inc,
2000 .

 Kamil Wiśniewski, July 12th, 2007, Discourse Analysis. Retrieved from:

http://www.tlumaczenia-angielski.info/linguistics/discourse.htm, Retrieved date: Oct
16, 2010.

 M.A.Khan, “ Text Based Machine Translation System”, PhD Thesis, 1995.

 The Daily News, “Jolie was high on cocaine during TV interview: Former drug
dealer”, dated: 22 Oct, 2010.
http://dailymailnews.com/1010/22/ShowBiz/index.php?id=3

 M.A.Khan, “MACHINE TRANSLATION BEYOND SENTENCE BOUNDARIES ”,, In Proceedings

of Workshop on Proofing Tools and Language Technologies, July 1-2, Patras University,
Greece.
www.mabidkhan.com/.../Scientific%20Khyber,%20Vol%201,%202004.pdf