Answer for exam

1) Major activities in NLP and what each does [5 pts]

Natural Language Processing involves several main activities:
1. Lexical analysis
 Deals with words and their meanings.
 Includes tokenization, stemming, and stop-word removal.
Example: Splitting a sentence into words.
2. Morphological analysis
 Studies internal structure of words (prefix, root, suffix).
Example: “unhappiness” → un + happy + ness.
3. Syntactic analysis (Parsing)
 Checks grammatical structure of sentences.
 Determines relationship among words.
Example: Identifying subject–verb–object.
4. Semantic analysis
 Extracts meaning from text.
Example: Understanding that “bank” may mean financial institution or river bank.
5. Discourse analysis
 Studies relationships between sentences in a paragraph.
Example: Resolving references like “he”, “she”, “it”.
6. Pragmatic analysis
 Considers context and intended meaning.
Example: “Can you open the window?” → request, not a question.
2) Two applications of NLP and how NLP helps [3 pts]
1. Machine Translation
 Converts text from one language to another.
Example: English → Amharic translation.
NLP helps by understanding grammar, vocabulary, and context.
2. Sentiment Analysis
 Determines emotion/opinion in text.
Example: Product review classification (positive/negative).
NLP extracts opinion words and context to predict sentiment.
Other examples: chatbots, speech recognition, text summarization.
3) Morphology components [5 pts]
a) Difference between morphological analysis and morphological generation [3 pts]
Morphological Analysis Morphological Generation
Breaks words into morphemes Builds words from morphemes
Input: word Input: root + grammar
Output: root, prefix, suffix Output: full word
Example:
 Analysis: “playing” → play + ing
 Generation: play + ing → playing
b) Advantages of morphological analysis [2 pts]
1. Helps understand word meaning
2. Improves machine translation accuracy
3. Helps in information retrieval
 4. Supports spelling correction
4) Techniques used in syntactic analysis and how they work [5 pts]
1. Part-of-Speech (POS) tagging
 Assigns grammatical category to words.
Example: Noun, verb, adjective.
2. Parsing
 Builds sentence structure tree.
Example: Subject–Verb–Object identification.
3. Context-Free Grammar (CFG)
 Uses grammar rules to validate sentences.
4. Dependency parsing
 Shows relationships between words.
How they work:
 Identify word types
 Apply grammar rules
 Build sentence structure
 Detect grammatical errors
5) Semantic analysis [12 pts]
a) Challenges in semantic analysis [3 pts]
1. Word ambiguity (polysemy)
2. Context dependency
3. Synonyms and homonyms
4. Idioms and figurative language
5. Lack of world knowledge
b) Methods to represent meaning in a sentence [3 pts]
1. Semantic networks
2. Ontologies
3. First-order logic
4. Vector space models
5. Word embeddings
c) Two synonymous and polysemy Amharic words [3 pts]
Synonyms:
 “ቆንጆ” and “ውብ” → both mean beautiful
 “ፈጣን” and “ፈጣሪ” (contextually speed-related usage)
Polysemy:
 “ባንክ”
o financial institution
o river bank
 “እጅ”
o hand
o help/support
d) Latent Semantic Analysis (LSA) and difference from topic modeling [3 pts]
Latent Semantic Analysis (LSA)
 Technique that identifies hidden relationships between words and documents.
 Uses matrix decomposition (SVD).
 Captures similarity between texts.
Topic Modeling
 Discovers topics present in documents.
 Example: politics, sports, business.
 Uses probabilistic models like LDA.
Difference:
LSA Topic Modeling
Based on linear algebra Based on probability
Finds word similarity Finds document topics
Uses SVD Uses LDA

6) Discourse and its focus in text processing [3 pts]

Discourse
Study of how sentences connect to form meaningful text.
Focus:
 Sentence relationships
 Context understanding
 Coherence and cohesion
Challenges:
1. Reference resolution (“he”, “it”)
2. Context interpretation
3. Ambiguity
Handling methods:
 Coreference resolution
 Context modeling
 Discourse parsing
7) Ambiguity in NLP with examples [7 pts]
Ambiguity occurs when a word or sentence has multiple meanings.
Types with examples:
1. Lexical ambiguity
Word has multiple meanings.
Example: “bank”
 river bank
 financial bank
2. Syntactic ambiguity
Sentence structure unclear.
Example:
“I saw a man with a telescope.”
 I used telescope
 Man had telescope
3. Semantic ambiguity
Meaning depends on context.
Example:
“Cold drink”
 chilled drink
 drink for illness
4. Pragmatic ambiguity
Depends on situation.
Example:
“Can you pass the salt?”
 ability question
 request

Additional for exam preparation

1) NLP Stages
Explain the major stages of Natural Language Processing with examples.
NLP stages are sequential processing steps used to analyze and understand human language by
computers, starting from word-level processing to context-level interpretation.
Main stages and explanation
1. Lexical analysis
o Breaks text into words/tokens.
o Removes punctuation and stop words.
Example:
“I love NLP” → [I, love, NLP]
2. Morphological analysis
o Studies structure of words (prefix, root, suffix).
Example:
“unhappiness” → un + happy + ness
3. Syntactic analysis
o Examines grammatical structure of sentences.
Example:
Identifies subject–verb–object.
4. Semantic analysis
o Extracts meaning from text.
Example:
“bank” → financial institution or river edge.
5. Discourse analysis
o Connects multiple sentences.
Example:
“John went home. He was tired.” (“He” refers to John)
6. Pragmatic analysis
o Understands intended meaning using context.
Example:
“Can you open the door?” → request, not question.
2) Applications of NLP
List and explain applications of NLP with examples.
NLP applications are real-world systems that use language processing to solve problems and
automate tasks involving human communication.
Major applications
1. Machine translation
o Converts text between languages.
Example: English → Amharic translation.
2. Sentiment analysis
o Detects emotion/opinion in text.
Example:
“This phone is amazing” → Positive.
3. Chatbots
o Automated conversation systems.
4. Speech recognition
o Converts speech into text.
5. Text summarization
o Produces short summaries of documents.
3) Morphology
Explain morphology and its two main components with examples.
Morphology is the study of the internal structure of words and how words are formed using
morphemes (smallest units of meaning).
Components
1. Morphological analysis
o Breaks words into parts.
Example:
“playing” → play + ing
2. Morphological generation
o Creates words from morphemes.
Example:
play + ed → played
Importance
 Helps translation
 Improves spell checking
 Helps text search
4) Syntax Techniques
Describe syntactic analysis and techniques used in it.
Syntax analysis studies sentence structure and grammatical relationships among words.
Techniques
1. POS tagging
o Assigns grammatical categories.
Example:
“She runs fast” → She(noun), runs(verb), fast(adverb)
2. Parsing
o Builds tree structure of sentence.
3. Context-Free Grammar
o Uses grammar rules.
4. Dependency parsing
o Shows relationships between words.
Example
Sentence: “The boy eats food”
 Subject → boy
 Verb → eats
 Object → food
5) Semantic Analysis
Define semantic analysis and explain its methods and challenges.
Semantic analysis is the process of understanding the meaning of words, phrases, and sentences.
Methods
 Semantic networks
 Ontologies
 Word embeddings
 Vector space models
Challenges
 Word ambiguity
 Context understanding
 Synonyms
 Idioms
Example
“Apple”
 fruit
 technology company
6) Discourse
What is discourse analysis and why is it important?
Discourse analysis studies how multiple sentences connect to form meaningful communication.
Focus
 Context
 Sentence relationships
 Coherence
Challenges
 Reference resolution
 Topic tracking
Example
“Mary bought a car. It is red.”
“It” refers to the car.
7) Ambiguity
Explain ambiguity in NLP and give examples.
Ambiguity occurs when a word, phrase, or sentence has more than one possible meaning.
Types
1. Lexical ambiguity
Word has multiple meanings.
Example: bank
2. Syntactic ambiguity
Sentence structure unclear.
Example:
“I saw a man with a telescope.”
 3. Semantic ambiguity
Meaning depends on context.
4. Pragmatic ambiguity
Depends on speaker intention.
Compact exam revision summary
Topic Core idea
NLP stages Steps to process language
Applications Real-world uses
Morphology Word structure
Syntax Sentence grammar
Semantics Meaning extraction
Discourse Sentence connection
Ambiguity Multiple meanings