CS-703-B Data Mining and Warehousing - 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
   aDiagrammatically illustrate and discuss the three tier data warehousing architecture.
   How to answer

   • Define three-tier warehouse architecture with bottom, middle and top tiers
   • Explain warehouse database server, OLAP server and client tools layers
   • Draw labelled block diagram showing data sources to clients
   • Discuss data flow, ETL and query processing
2. 7 Marks
   bDiscuss the following:

   1. i)Star schema
   2. ii)Snow Flake schema
   3. iii)Fact constellation schema
   How to answer

   • Define star schema with central fact table and denormalized dimensions
   • Define snowflake schema with normalized dimension hierarchies
   • Define fact constellation with multiple fact tables sharing dimensions
   • Compare structure, redundancy and query performance
3. 7 Marks
   aBriefly discuss the schemas for multidimensional databases.
   How to answer

   • Define multidimensional databases and need for schemas
   • Explain star, snowflake and fact constellation schemas briefly
   • Explain fact and dimension tables, measures and hierarchies
   • Give example for sales or similar warehouse
4. 7 Marks
   bGive detailed information about Meta data in data warehousing.
   How to answer

   • Define metadata and its role in warehousing
   • Explain operational, extraction and end-user metadata
   • Describe metadata repository and management
   • Explain uses in ETL, querying and administration
5. 7 Marks
   aExplain the categorization of OLAP tools with necessary diagrams.
   How to answer

   • Define OLAP and need for categorization
   • Explain ROLAP, MOLAP and HOLAP with architecture
   • Compare storage, performance and scalability
   • Draw necessary diagrams for each category
6. 7 Marks
   bList and explain the OLAP operation in multidimensional data model.
   How to answer

   • Define multidimensional data model with cube, dimensions and measures
   • List roll-up, drill-down, slice, dice and pivot operations
   • Explain each operation with example
   • Show effect on data cube
7. 7 Marks
   aHow a data mining system can be integrated with a data warehousing? Discuss with example.
   How to answer

   • Describe coupling architectures: no coupling, loose, semi-tight, tight
   • Explain integration of mining with warehouse server/OLAP
   • Give example of mining sales warehouse data
   • Discuss advantages of warehouse for mining
8. 7 Marks
   bExplain with diagrammatic illustration data mining as a step in the process of knowledge discovery.
   How to answer

   • Define KDD process steps from selection to presentation
   • Locate data mining as central pattern extraction step
   • Draw KDD process diagram with iterative loops
   • Explain preprocessing, mining and evaluation roles
9. 7 Marks
   aWrite short notes on the various preprocessing tasks.
   How to answer

   • Define preprocessing and its need
   • Explain cleaning, integration, transformation and reduction
   • Give techniques for each task
   • Summarize impact on mining quality
10. 7 Marks
    bExplain data mining vs knowledge discovery in databases.
    How to answer

    • Define data mining vs KDD
    • Compare scope, steps and output
    • Explain mining as one step of KDD
    • Give example workflow
11. 7 Marks
    aExplain whether association rule mining is supervised or unsupervised type of learning.
    How to answer

    • Define supervised vs unsupervised learning
    • Explain association rule mining with support and confidence
    • Argue no labeled target, hence unsupervised/descriptive
    • Give market basket example
12. 7 Marks
    bClassification is supervised learning? Justify.
    How to answer

    • Define classification and supervised learning
    • Explain training with class labels
    • Explain model construction and testing phases
    • Justify with example like decision tree
13. 7 Marks
    aDefine Clustering. Explain about types of data in cluster analysis.
    How to answer

    • Define clustering as unsupervised grouping
    • Explain types: interval, binary, categorical, ordinal, mixed
    • Explain distance measures for each type
    • Give examples
14. 7 Marks
    bWhat is the goal of clustering? How does partitioning around medoids algorithm achieve this?
    How to answer

    • State goal: maximize intra-cluster similarity, minimize inter-cluster similarity
    • Explain PAM/PAM partitioning around medoids concept
    • Outline swap-based optimization steps
    • Explain robustness and cost
15. 7 Marks
    aExplain the following clustering methods in detail.

    1. i)BIRCH
    2. ii)CURE
    How to answer

    • Explain BIRCH with CF-tree and hierarchical clustering
    • Explain phases: building tree, condensing and clustering
    • Explain CURE with representative points and shrinking
    • Highlight handling of large data and arbitrary shapes
16. 7 Marks
    bExplain about the Apriori algorithm for finding frequent item sets with an example.
    How to answer

    • Define frequent itemsets, support and confidence
    • Explain Apriori property and join-prune steps
    • Illustrate iterative candidate generation with example
    • Show final association rules generation