AL-504-B Natural Language Processing - Nov 2022 Question Paper

Under each question, How to answer lists the points an answer should cover, outlined by CampusPrep from the syllabus. Check the details against your textbook.

1. 7 Marks
   aList and explain the challenges of natural language processing.
   How to answer

   • Define NLP and need for challenges
   • List challenges: ambiguity, variability, context, etc.
   • Brief explanation of each challenge with example
2. 7 Marks
   bDiscuss in detail the basic regular expression patterns with examples.
   How to answer

   • Define regular expression in NLP
   • List basic patterns: concatenation, disjunction, Kleene star, etc.
   • Give example for each pattern
   • Mention use in text search and tokenization
3. 7 Marks
   aWrite and explain the minimum edit distance algorithm.
   How to answer

   • Define minimum edit distance and operations
   • Present dynamic programming recurrence
   • Explain algorithm steps with matrix filling
   • Give example trace
4. 7 Marks
   bExplain with an example how bigram probability is calculated.
   How to answer

   • Define bigram and bigram probability formula
   • Explain MLE counting: C(w1 w2)/C(w1)
   • Work out numeric example on sample corpus
   • Mention use in language modeling
5. 7 Marks
   aExplain how is perplexity calculated in evaluating N-grams.
   How to answer

   • Define perplexity as evaluation metric
   • Give perplexity formula for N-grams
   • Explain steps to compute on test set
   • Interpret lower vs higher perplexity
6. 7 Marks
   bDiscuss in detail the Hidden Markov Model.
   How to answer

   • Define Hidden Markov Model and hidden vs observed states
   • Explain components: transition, emission, initial probabilities
   • Describe three problems: evaluation, decoding, learning
   • Mention application to POS tagging
7. 7 Marks
   aWhat is ambiguity in parse tree? Explain with an example.
   How to answer

   • Define parse tree and ambiguity
   • Explain types with example sentence
   • Show two different parse trees for same sentence
   • Mention impact on parsing
8. 7 Marks
   bExplain in detail about dependency grammar.
   How to answer

   • Define dependency grammar vs constituency
   • Explain head-dependent relations
   • Describe dependency tree structure
   • Give example sentence analysis
9. 7 Marks
   aDiscuss in detail Probabilistic CYK.
   How to answer

   • Define Probabilistic CYK and CNF requirement
   • Explain dynamic programming table with probabilities
   • Describe steps to find most probable parse
   • Mention application in PCFG parsing
10. 7 Marks
    bExplain in detail about quantifiers with an example.
    How to answer

    • Define quantifiers in semantic representation
    • Explain universal and existential quantifiers
    • Give logical form example
    • Mention scope and ambiguity
11. 14 Marks
    Explain the following word sense disambiguation methods:

    1. a)Supervised
    2. b)Dictionary
    3. c)Thesaurus
    How to answer

    • Define word sense disambiguation
    • Explain supervised method with training data
    • Explain dictionary/Lesk method
    • Explain thesaurus-based method
12. 7 Marks
    aDescribe transfer model of machine translation. List out its three phases.
    How to answer

    • Define machine translation transfer model
    • Describe analysis, transfer, generation phases
    • List three phases explicitly
    • Give example of transfer rules
13. 7 Marks
    bWhat are the commercial uses of NLP? Explain in detail.
    How to answer

    • Define commercial applications of NLP
    • List uses: search, translation, chatbots, sentiment analysis
    • Explain each use in detail
    • Give real-world examples
14. 14 Marks
    Write short notes on any two of the following:

    1. a)Finite-State Automata
    2. b)Laplace Smoothing
    3. c)Probabilistic context-free grammar
    4. d)Word sense
    How to answer

    • Select any two topics from list
    • Define each selected concept clearly
    • Explain key formulas/steps with example
    • Mention applications in NLP