AL-504-B Natural Language Processing - Dec 2024 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
   aDefine Natural Language Processing, what are the key components and challenges of it.
   How to answer

   • Define Natural Language Processing as machine processing of human language
   • List key components e.g. morphological, syntactic, semantic, pragmatic analysis
   • List challenges e.g. ambiguity, polysemy, multilinguality, context
   • Give examples of applications
2. 7 Marks
   bDiscuss finite-state automata and their relationship to regular expressions.
   How to answer

   • Define finite-state automata and regular expressions
   • Explain equivalence between FSA and regular languages
   • Describe use in lexical analysis and morphology
   • Give examples
3. 7 Marks
   aExplain the morphological analysis using stemming and lemmatization processes.
   How to answer

   • Define morphological analysis
   • Explain stemming with examples
   • Explain lemmatization with examples
   • Compare stemming vs lemmatization
4. 7 Marks
   bHow do the modeling approaches of Hidden Markov Models and Maximum Entropy differ in terms of probabilistic dependencies?
   How to answer

   • Define Hidden Markov Models as generative directed model
   • Define Maximum Entropy / MaxEnt as discriminative log-linear model
   • Contrast joint vs conditional dependencies and independence assumptions
   • Note implications for feature use and sequence labeling
5. 7 Marks
   aCompare common types of smoothing techniques used in natural language processing.
   How to answer

   • Define smoothing purpose for zero-probability n-grams
   • Describe Laplace, Good-Turing, interpolation, backoff, Kneser-Ney types
   • Compare principle and use-case of each
   • Summarize advantages and limitations
6. 7 Marks
   bWhy are n-gram models limited in their ability to handle polysemy and ambiguity in language?
   How to answer

   • Define n-gram models and fixed context window
   • Explain lack of long-range context and semantics
   • Explain polysemy and ambiguity handling failure
   • Give examples
7. 7 Marks
   aDescribe disambiguation techniques and challenges in parsing.
   How to answer

   • Define parsing ambiguity types
   • Describe disambiguation techniques e.g. probabilistic, grammar-based
   • List challenges e.g. PP-attachment, coordination
   • Give examples
8. 7 Marks
   bCheck the acceptance of string $s="baaba"$ using the CYK Algorithm for the given grammar $S \rightarrow AB / BC$ $A \rightarrow BA / a$ $B \rightarrow CC / b$ $C \rightarrow AB / a$
   How to answer

   • Convert grammar to Chomsky Normal Form check and list productions
   • Build CYK table bottom-up for s = baaba
   • Show cell derivations for substrings
   • Conclude acceptance if S in top cell
9. 7 Marks
   aExplain how dependency grammar differs from phrase-structure grammar.
   How to answer

   • Define dependency grammar with head-dependent relations
   • Define phrase-structure grammar with constituency
   • Contrast representation, hierarchy and word order handling
   • Give example trees
10. 7 Marks
    bIllustrate Word similarity using a thesaurus and distributional methods.
    How to answer

    • Define word similarity
    • Explain thesaurus-based similarity e.g. WordNet path
    • Explain distributional / vector similarity
    • Illustrate with examples
11. 7 Marks
    aBriefly explain different Bootstrapping methods in NLP.
    How to answer

    • Define bootstrapping for low-resource learning
    • Describe methods e.g. self-training, co-training, Yarowsky
    • Explain iterative seed expansion steps
    • Note applications
12. 7 Marks
    bExplain supervised Word Sense Disambiguation methods with suitable examples.
    How to answer

    • Define Word Sense Disambiguation
    • Explain supervised approaches e.g. Naive Bayes, decision lists, SVM
    • Describe training with sense-annotated corpora and features
    • Give suitable examples
13. 7 Marks
    aHow can natural language processing automate the authoring process? Discuss with applications.
    How to answer

    • Define authoring automation in NLP
    • Describe techniques e.g. NLG, summarization, templates
    • Discuss applications e.g. report generation, content creation
    • Note benefits
14. 7 Marks
    bWhat makes natural language querying possible through natural language processing?
    How to answer

    • Define natural language querying / NLIDB
    • Explain pipeline: parsing, semantic mapping, SQL/query generation
    • Discuss role of lexicons, grammars and dialogue
    • Give examples
15. 14 Marks
    Explain any two of the following

    1. a)Shallow parsing
    2. b)Transducers for lexicon and rules
    3. c)Dictionary and Thesaurus
    4. d)Man-Machine interfaces
    How to answer

    • Select any two topics from a-d
    • Define and explain each with key concepts
    • Give examples and applications
    • Add diagram or steps where relevant