Natural Language Processing

By

​

S.V.V.D.Jagadeesh

Sr. Assistant Professor

Dept of Artificial Intelligence & Data Science

​

​

LAKIREDDY BALI REDDY COLLEGE OF ENGINEERING

• Previously Discussed Topics
• Session Outcomes
• Description Logic
• Components of DL
• Concepts (Classes)
• Roles (Relationships)
• Individuals (Instances)
• DL Knowledge Base
• Terminological Box (TBOX)

​

​

​

​

​

​

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Previously Discussed Topics

LBRCE

NLP

• Assertional Knowledge (ABox)
• Concept Constructors in DL
• Conjunction (AND)
• Disjunction (OR)
• Negation (NOT)
• Existential Restriction
• Universal Restriction
• Exercises
• Solutions
• Advantages of DL
• Disadvantages of DL

​

​

​

​

​

​

​

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Previously Discussed Topics

LBRCE

NLP

At the end of this session, Student will be able to:

• Understand Syntax-Driven semantic Analysis and Semantic Attachments(Understand-L2)

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Session Outcomes

LBRCE

NLP

• Syntax-driven semantic analysis is the method of computing the meaning of a sentence based on its syntactic structure.
• The syntactic parse tree guides the semantic interpretation, and semantic rules are attached to grammar rules.
• Thus:
• Meaning of a sentence = Meaning of its parts + Syntactic structure
• This approach is widely used in semantic parsing and compositional semantics.

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Syntax-Driven Semantic Analysis

LBRCE

NLP

• The Principle of Compositionality states:
• The meaning of a sentence is determined by the meanings of its individual words and the way they are syntactically combined.
• This principle is fundamental in syntax-driven semantic analysis.
• Example
• Sentence: The dog chased the cat
• Word meanings: dog cat chase
• Semantic composition: chase(dog, cat)
• The meaning is obtained by combining the meanings of the words using syntactic structure.

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Principle of Compositionability

LBRCE

NLP

• Context-Free Grammar (CFG) describes sentence structure.
• Example rule: VP → V NP
• To build meaning, CFG rules are augmented with semantic rules.
• Example
• Grammar rule: VP → V NP
• Semantic rule: VP.sem = V.sem(NP.sem)
• Meaning::The verb meaning is applied to the noun phrase meaning.
• Example Sentence: Mary likes chocolate
• Parse: VP → likes chocolate
• Semantic representation: like(Mary, chocolate)

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Semantic Augmentations to CFG Rules

LBRCE

NLP

• Sentences with quantifiers may have multiple interpretations depending on scope.
• Example sentence: Every student read a book
• Two possible meanings:
• Interpretation 1: Each student read possibly different books.
• ∀x Student(x) → ∃y Book(y) ∧ Read(x,y)
• Interpretation 2: There is one book that all students read.
• ∃y Book(y) ∧ ∀x Student(x) → Read(x,y)
• Both interpretations are grammatically valid.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Quantifier Scope Ambiguity

LBRCE

NLP

• A unification-based approach combines syntactic and semantic features using feature structures.
• These feature structures are unified during parsing.
• Feature Structure Example
• [ CAT = NP� SEM = John ]
• Example Grammar Rule: S → NP VP
• Semantic unification: S.sem = VP.sem(NP.sem)
• Example Sentence: John runs
• Unification result: run(John)

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Unification Based Approaches

LBRCE

NLP

• Semantic attachments associate semantic interpretation rules with grammar rules.
• These rules allow computing meaning during parsing.
• We illustrate semantic attachments for common phrase types.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Semantic Attachments

LBRCE

NLP

• Grammar rule
• S → NP VP
• Semantic rule
• S.sem = VP.sem(NP.sem)
• Example
• John sleeps
• Representation
• sleep(John)

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Sentences

LBRCE

NLP

• Noun Phrase
• Example: the dog
• Semantic representation: dog(x)
• Genitive Noun Phrase
• Example: John's book
• Semantic Representation: book(y) ∧ possess(John,y)
• Meaning: The book belongs to John.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Noun and Genitive Noun Phrases

LBRCE

NLP

• Adjectives modify nouns.
• Example: big dog
• Semantic representation: dog(x) ∧ big(x)
• Example
• red car
• Representation: car(x) ∧ red(x)

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Adjective Phrases

LBRCE

NLP

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Verb Phrases

• Verb phrases represent actions or events.
• Example: Mary likes chocolate
• Representation: like(Mary, chocolate)
• Roles:

​

LBRCE

NLP

• An infinitive verb phrase contains the word to.
• Example: John wants to eat pizza
• Representation: want(John, eat(John,pizza))
• Meaning: John desires the action of eating pizza.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Infinite Verb Phrases

LBRCE

NLP

• Prepositional phrases express relationships such as location, instrument, or time.
• Example: The book is on the table
• Representation: on(book, table)

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Prepositional Phrases

LBRCE

NLP

• These modify nouns.
• Example: the book on the table
• Meaning: book(x) ∧ on(x, table)
• The phrase on the table modifies book.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Nominal Modifier Prepositional Phrases

LBRCE

NLP

• These modify verbs or actions.
• Example: John ran in the park
• Representation: run(John) ∧ location(run, park)
• Meaning: The action occurred in the park.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Verb Phrase Modifier Prepositional Phrases

LBRCE

NLP

• Some verbs require prepositional phrases as arguments.
• Example: John put the book on the table
• Representation: put(John, book, table)
• Here: on the table is a required argument of the verb put.

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Verb Argument Prepositional Phrases

LBRCE

NLP

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Idioms and Compositionability

• Idioms are expressions whose meanings cannot be derived from individual word meanings.
• Example
• kick the bucket
• Literal meaning
• kick(bucket)
• Actual meaning
• die

​

LBRCE

NLP

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Challenges with Idioms

• They violate the principle of compositionality.
• Meaning must be stored as a single lexical unit.
• Example
• Sentence
• He kicked the bucket
• Interpretation
• die(he)

​

LBRCE

NLP

• Previously Discussed Topics
• Session Outcomes
• Syntax- Driven Semantic Analyis
• Principle of Compositionability
• Semantic Augmentation to CFG Rules
• Quantifier Scope Ambiguity
• Unification Based Approaches
• Semantic Attachments
• Sentences

​

​

​

​

​

​

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Summary

LBRCE

NLP

• Noun and Genitive Noun Phrases
• Adjective Phrases
• Verb Phrases
• Infinite Verb Phrases
• Prepositional Phrases
• Nominal modifier Prepositional Phrases
• Verb Phrase Prepositional Phrases
• Verb Argument Prepositional Phrases
• Idioms and Compositionability
• Challenges with Idioms

​

​

​

​

​

​

​

​

S.V.V.D.Jagadeesh

Wednesday, March 11, 2026

Summary

LBRCE

NLP