Preview Test: DATA7201 Semester One Final Examination 2020
Test Information
Description
Instructions
Timed Test This test has a time limit of 2 hours and 30 minutes.This test will save and be submitted
automatically when the time expires.
Warnings appear when half the time, 5 minutes, 1 minute, and 30 seconds remain.
[The timer does not appear when previewing this test]
Multiple
Attempts
Not allowed. This Test can only be taken once.
Force
Completion
This Test can be saved and resumed at any point until time has expired. The timer will
continue to run if you leave the test.
Your answers are saved automatically.
There are 5 open questions and 2 scenario discussions.
Each of the 5 questions is worth 14 points. Each of the 2 scenarios is worth 15 points.
Answer each question and discuss the scenarios describing the type of data
infrastructure you would adopt and why. The scenario descriptions below do not
provide all the necessary details. You will need to describe your assumptions on the
scenarios to complement the information given to you for each scenario. Describe the
assumptions you are making in terms of data availability, analytics queries of
interest, user expertise and requirements. For each scenario, 1) discuss your
assumptions, 2) outline the design of your data infrastructure solution (i.e., which
data, which systems, which users, etc.) and, 3) justify your solution.
QUESTION 1
Explain the role and the functioning of the Primary NameNode in
HDFS.
14 points   Save Answer
QUESTION 2
Explain the challenges in Hadoop Map/Reduce job scheduling
for sequential Map/Reduce jobs (i.e., for jobs that require to be
executed strictly one after the other).
14 points   Save Answer
QUESTION 3
Critically compare the functioning of Apache Kafka and Spark
Streaming.
14 points   Save Answer
QUESTION 4
Explain possible ways of executing PageRank over a large dataset in a
distributed environment. Discuss the key challenges involved in doing
it
14 points   Save Answer
QUESTION 5
Discuss the main disadvantages of Collaborative Filtering
approaches for recommender systems.
14 points   Save Answer
QUESTION 6
A large company that manages a chain of shopping malls at different
geographical locations wants to collect and analyse data to understand
trends of product sales. The results of the data analysis will need to be
included in a report used to inform strategic decisions for the company
board meetings happening monthly.
15 points   Save Answer
QUESTION 7
A company is running a massive multiplayer online game and wants
to process real-time trading information from the game (i.e., players
trading game items for virtual gold or other game credits) to identify
suspicious events that could indicate players cheating by using
scripts and other tools to gain more items or credits in the game.
15 points   Save Answer
QUESTION 8
Free-text zero mark question for student write notes
0 points   Save Answer