AL-503-A Information Retrieval - Nov 2023 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
   aExplain the Information retrieval process with key components.
   How to answer

   • Definition of Information Retrieval and its goal
   • Steps of IR process: document collection, preprocessing, indexing, query processing, matching/ranking, evaluation
   • Key components: crawler/collection, indexer, query processor, retrieval model, user interface
   • Diagram of IR process flow
2. 7 Marks
   bDiscuss the role of Artificial Intelligence (AI) in IR.
   How to answer

   • Role of AI: NLP, ML, deep learning in IR
   • Examples: semantic search, ranking, recommendation, QA
   • Benefits: relevance, personalization, automation
   • Challenges briefly
3. 7 Marks
   aContrast Traditional IR systems with web search.
   How to answer

   • Characteristics of traditional IR: closed corpus, structured documents, Boolean/vector models
   • Characteristics of web search: scale, heterogeneity, link structure, dynamic
   • Comparison on corpus, queries, ranking, evaluation
   • Summary table
4. 7 Marks
   bWrite short notes on the probabilistic relevance feedback.
   How to answer

   • Definition of relevance feedback and probabilistic view
   • Probabilistic relevance feedback formula/idea
   • Iterative query expansion process
   • Example and advantages/limitations
5. 7 Marks
   aBriefly explain the Document Preprocessing procedure.
   How to answer

   • Purpose of preprocessing in IR
   • Steps: tokenization, stop-word removal, stemming/lemmatization, normalization
   • Example of transformation
   • Effect on index size and retrieval
6. 7 Marks
   bExplain latent semantic indexing with an example.
   How to answer

   • Concept of latent semantic indexing and SVD
   • Term-document matrix construction
   • Dimensionality reduction example with small matrix
   • Use for synonymy/polysemy handling
7. 7 Marks
   aState the advantages and needs of a crawler.
   How to answer

   • Definition of web crawler
   • Needs: collection building, freshness, coverage
   • Advantages: automation, scalability, indexing support
   • Brief working
8. 7 Marks
   bExplain the Benefits of index compression with an example.
   How to answer

   • Why index compression is needed: space and speed
   • Techniques: dictionary compression, postings compression e.g. gamma/delta codes
   • Worked example of postings list compression
   • Benefits in storage and query performance
9. 7 Marks
   aDescribe the requirements of the XML information retrieval system.
   How to answer

   • XML IR vs flat document retrieval
   • Requirements: structure-aware indexing, CAS/NEXI queries, relevance of elements
   • Challenges: nested elements, overlap, presentation
   • Evaluation aspects
10. 7 Marks
    bAnalyze the effectiveness of a MapReduce and Hadoop implementation.
    How to answer

    • Basics of MapReduce and Hadoop architecture
    • Use for large-scale indexing and retrieval
    • Effectiveness measures: scalability, fault tolerance, speed
    • Limitations/trade-offs
11. 7 Marks
    aExplain in detail the Content-based recommendation system.
    How to answer

    • Principle of content-based filtering
    • Item profile and user profile construction
    • Similarity computation and recommendation steps
    • Example plus pros/cons
12. 7 Marks
    bDescribe the retrieval procedure using the ranking and relative score.
    How to answer

    • Concept of ranking and relative score
    • Scoring functions: TF-IDF, cosine, BM25
    • Steps from query matching to ranked list
    • Example of score calculation
13. 7 Marks
    aExplain the process of choosing K in Nearest neighbor clustering.
    How to answer

    • KNN clustering concept and role of K
    • Effect of small vs large K
    • Methods: elbow, cross-validation, heuristic
    • Example selection process
14. 7 Marks
    bDiscuss in detail the Naïve Bayesian classification and its limitation.
    How to answer

    • Bayes theorem and naive independence assumption
    • Classification steps with example calculation
    • Advantages: simplicity, efficiency
    • Limitations: independence, zero probability, skewed data
15. 14 Marks
    Explain the following (any two)

    1. a)Snippet Generation
    2. b)Agglomerative Clustering
    3. c)Focused Crawling
    4. d)TF-IDF term weighting
    How to answer

    • For each chosen topic: definition and purpose
    • Working steps/algorithm
    • Example or application in IR
    • Advantages/limitations