Name
Reg No B
APJ ABDUL KALAM TECHNOLOGICAL UNIVERSITY
07 THRISSUR CLUSTER
THIRD SEMESTER M.TECH. DEGREE EXAMINATION DEC 2017
COMPUTER SCIENCE DEPARTMENT
COMPUTER SCIENCE AND ENGINEERING
07CS7115 BIG DATA ANALYTICS
Time:3 hours Max.marks: 60
Answer all six questions. Part of each question is compulsory.
Answer either part or part of each question
Q.no. Module 1 Marks
1a. What are the different data structures supported by Big Data? Give
examples?
4
Answer b or c
b. Evolution of big data brings with it new security challenges. Give your
view.
5
c. Is EDW a good choice for data repository from an analyst perspective?
Give reasons to support your answer.
5
Q.no. Module 2 Marks
2a 'Data conditioning is a data pre-processing step'. Comment. 4
Answer b or c
b The analytics team needs to envision the risks involved in doing the
analysis project. Suggest the phase of data analytics life cycle which is
suitable for this operation.
5
c Data analytic team found out that neural network is a good technique for
meeting the business requirement. Give a detailed analysis of the data
analytic phase they are undergoing, highlighting the useful tools.
5
Q.no. Module 3 Marks
3a What are the NOIR Attributes in 4
Answer b or c
b Consider an analysis context to test how weight of a person can be related
to health problems. There are 4 categories of people, namely underweight,
normal weight, overweight and obese. Which statistical method is suitable
for the situation? Explain with support of R programming language.
5
c Give a comparative study on descriptive statistical analysis and exploratory
data analysis, with example.
5
Q.no. Module 4 Marks
4a What are the components of a time series? Explain. 4
Answer b or c
b Is it necessary to use ACF in certain time series for further processing?
Give your view.
5
c For an model, autocorrelations have a value of zero beyond lag
Give reasons.
5
Q.no. Module 5 Marks
5a What are the different approaches to represent text during analysis? 5
Answer b or c
b Consider a scenario where a company stopped manufacturing a product
based on the reviews it received from social networking sites. What kind of
analysis would they have applied to make such a decision?
7
c Large collection of unstructured text data in Tera byte size need to be
analysed in short time. Suggest a suitable method for performing it in a time
bound fashion.
7
Q.no. Module 6 Marks
6a What is a NoSQL data store? Explain its different types. 5
Answer b or c
b Suppose a website wants to store its real-time data feeds. Which Hadoop
technology can be used for this scenario? Give reasons.
7
c Suppose that finding Word count from a terra byte size of data is an urgent
requirement. Suggest a technology that can handle this requirement
7