Syntactic and Semantic Analysis

Syntactic Analysis (Parsing)

Syntax concerns how words combine to form grammatically correct sentences. Syntactic analysis checks a sentence against the grammar rules of a language and produces a structural representation, usually a parse tree.

        S
      /   \
    NP      VP
    |      /  \
  "Riya"  V    NP
         |     |
       "reads" "books"

A grammar defines rules such as:

S  -> NP VP
NP -> Det N | N
VP -> V NP

import nltk
from nltk import CFG, ChartParser

grammar = CFG.fromstring("""
S -> NP VP
NP -> 'Riya' | 'books'
VP -> V NP
V -> 'reads'
""")
parser = ChartParser(grammar)
for tree in parser.parse(['Riya', 'reads', 'books']):
    print(tree)
# (S (NP Riya) (VP (V reads) (NP books)))

Syntactic ambiguity arises when a sentence has more than one valid parse tree:

# "I saw the man with a telescope" has two valid parses:
# 1. I used a telescope to see the man   (PP attaches to VP)
# 2. The man had a telescope             (PP attaches to NP "the man")

We study grammars and parsing algorithms (CFG, top-down, bottom-up, CYK) in detail in Unit 2.

Semantic Analysis

Semantics is the study of meaning. While syntax checks whether a sentence is grammatically well-formed, semantics checks whether it is meaningful.

Classic example (Noam Chomsky): "Colorless green ideas sleep furiously." This sentence is syntactically correct (NP + VP structure) but semantically anomalous — "colorless" and "green" contradict, and "ideas" cannot literally "sleep".

Semantic analysis involves:

# Semantic role labeling (conceptual)
sentence = "Riya gave Zoya a book."
# Agent:      Riya
# Recipient:  Zoya
# Theme:      a book
# Action:     gave

Semantic analysis is what lets a system understand that "The cat chased the mouse" and "The mouse was chased by the cat" describe the same event, despite different syntax (active vs passive voice).