CHINHOYI UNIVERSITY OF

TECHNOLOGY

Entrepreneurship & Business Sciences

Graduate Business School
MSc. Big Data Analytics

Big Data Structures and Algorithms

[MSCBDA 624]

Eng. N.F Thusabantu - Shoniwa

TOPIC 2 : Features of Big
Data
• Security

• Compliance

• Auditing and Protection

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 Content
1. Big Data Security Analytics
o Risk Management
o Visualizing Threat
o Penetration Test
2. Challenges to Big Data Security
3. Big Data Security Strategy
4. BDA use in Security analytics
o Network Traffic
o Web Transactions
o Network Server
o User Credentials
5. Adopting Security Analytics Solution
6. Compliance
7. Audit and Protection

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 Rapid7’s National
Exposure Index
• Rapid7’s report of the most
hackable countries for the
year 2017 puts Zimbabwe as
the most exposed country on
the earth – out of 183
countries analysed for the
report – with 54.71% of its
internet unencrypted.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 WHAT ARE WE
PROTECTING ?

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

1. Big Data Security Analytics:
A New Generation of Security Tools

• As the security industry’s response to these challenges, a new

generation of security analytics solutions has emerged in
recent years, which are able to collect, store and analyse huge
amounts of security data across the whole enterprise in real
time.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 a. Risk Management:
• It is essential to ensure your tools are supported by the
highly intelligent risk-management so that Big data can
be interpreted easily.

• In this way, you can categorize and handle the threats

properly without any delay.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 b. Visualize the Threat:
• Using Big data analytics, the analyst can predict the class
and intensity of the threat.

• With the help of data source evaluation, you can easily

judge the complexity of the cyber attack.

• The tools give the leverage to compare the historical and

current data to understand the stats of trend.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 c. Penetration Testing:
• It involves the five stages that are planning, scanning,
gaining, maintaining and analysis.

• Basically, it gives the insights of business infrastructure

i.e. database, a process so that hackers are not able to break
the security.

• Penetration testing is now an important step to protect the

business data and IT infrastructure.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

2. Challenges to Big Data
Security

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

a. Protecting Transaction Logs And
Data
• Data stored in a storage medium, such as transaction logs and
other sensitive information, may have varying levels, but
that’s not enough.
• For instance, the transfer of data between these levels gives
the IT manager insight over the data which is being moved.
• Data size being continuously increased, the scalability and
availability makes auto-tiering necessary for big data storage
management.
• New challenges are being posed to big data storage as the
auto-tiering method doesn’t keep track of data storage
location. 

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

b. Validation And Filtration Of
End-point Inputs
• End-point devices are the main factors for maintaining big
data.

• Storage, processing and other necessary tasks are performed

with the help of input data, which is provided by end-points.

• Therefore, an organization should make sure to use an

authentic and legitimate end-point devices.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 c. Securing Distributed
Framework Calculations And
Other Processes
• Computational security and other digital assets in a distributed
framework like MapReduce function of Hadoop, mostly
lack security protections.

• The two main preventions for it are securing the mappers

and protecting the data in the presence of an unauthorized
mapper.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 d. Securing And Protecting
Data In Real Time
• Due to large amounts of data generation, most  organizations
are unable to maintain regular checks.

• However, it is most beneficial to perform security checks and

observation in real time or almost in  real time.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 e. Protecting Access Control
Method Communication And
Encryption 
• A secured data storage device is an intelligent step in order
to protect the data.

• Yet, because most often data storage devices are vulnerable, it

is necessary to encrypt the access control methods as well.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 f. Data Provenance
• To classify data, it is necessary to be aware of its origin In
order to determine the data origin accurately, authentication,
validation and access control could be gained.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 g. Granular Auditing
• Analysing different kinds of logs could be advantageous and
this information could be helpful in recognizing any kind of
cyber attack or malicious activity.

• Therefore, regular auditing can be beneficial.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 h. Granular Access
Control

• Granular access control of big data stores by NoSQL databases

or the Hadoop Distributed File System requires a strong
authentication process and mandatory access control.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 i. Privacy Protection For
Non-relational Data
Stores
• Data stores such as NoSQL have many security vulnerabilities,
which cause privacy threats.

• A prominent security flaw is that it is unable to encrypt data

during the tagging or logging of data or while distributing it
into different groups, when it is streamed or collected.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 3. Big Data Security Strategy
• Big data security’s mission is clear enough: keep out on unauthorized users
and intrusions with firewalls, strong user authentication, end-user training,
and intrusion protection systems (IPS) and intrusion detection systems (IDS).
In case someone does gain access, encrypt your data in-transit and at-rest.

• This sounds like any network security strategy.

• However, big data environments add another level of security because

security tools must operate during three data stages that are not all present
in the network.

• These are
❖ 1. Data ingress (what’s coming in),
❖ 2. Stored data (what’s stored), and
❖ 3. Data output (what’s going out to applications and reports).
Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA
❖Stage 1: Data Sources. 
• Big data sources come from a variety of sources and data
types.

• User-generated data alone can include CRM or ERM data,

transactional and database data, and vast amounts
of unstructured data such as email messages or social media
posts.

• In addition to this, you have the whole world of machine

generated data including logs and sensors. You need to secure
this data in-transit from sources to the platform.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

❖Stage 2: Stored Data. 
• Protecting stored data takes mature security toolsets including
encryption at rest, strong user authentication, and
intrusion protection and planning.

• You will also need to run your security toolsets across a

distributed cluster platform with many servers and nodes.

• In addition, your security tools must protect log files

and analytics tools as they operate inside the platform.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

❖Stage 3: Output Data. 
• The entire reason for the complexity and expense of the big
data platform is being able to run meaningful analytics across
massive data volumes and different types of data.

• These analytics output results to applications, reports, and

dashboards.

• This extremely valuable intelligence makes for a rich target for

intrusion, and it is critical to encrypt output as well as ingress.

• Also, secure compliance at this stage: make certain that results

going out to end-users do not contain regulated data.
Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA
 Big Data: A Savior?
• Nowadays, most of the companies are depending on Big data
to develop the shield against cyber attacks. This helps the
cyber analyst in predicting the possibilities of invasion and
intrusion.

• Around 84% of the enterprises are using data to block the

hackers and because of that, there’s a decline in cybersecurity
breaches.

• In fact, the data can be used as an advanced analytics tool

which can give you the insights of cybersecurity threats
such as malicious insider programs, malware/ Ransomware
attacks, compromised and weak devices.
Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA
 Data Sight

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA
 4. BDA use in Security
Analytics
Detection and Prediction in:

• Network Traffic: suspicious traffic patterns.

• Web Transactions: abnormal user access patterns in the usage of critical resources
or activities.
❖ Identity theft.

• Network Servers: abnormal patterns related to server manipulation, e.g., abnormal

or sudden configuration changes
❖ Fraud detection.

• User Credentials: Detecting anomalies with respect to a user, or a group of user, not
complying with its inherent access behaviour, e.g., abnormal access time or
transaction amount.
❖ Card Cloning

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 Big Data Analytics
Forensics
• ChromePass demo
• GAIT Forensics used to solve murder
• RFI prediction (Remote File Inclusion)

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 5. Adopting Security
Analytics Solution
1. Develop Security Business Strategy
2. Participate in Analytics Training and Workshops
3. Implement a Centralized Data Management
Infrastructure
4. Implement an Analytics Platform
5. Hire Data Scientist as a Consultant
6. Implement a “Network Monitoring” Layer
7. Implement a “Suspicion Alert” Layer
8. Find out how to Streamline Analytics with Current
Workflow

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 6. Compliance
❖Compliance Definition
• Compliance can be understood as obliging in relation to
established guidelines, or the process of becoming so.
• For instance, Software is developed in compliance with
specifications constructed by a standards body, which is
then deployed by user organizations in acquiescence with a
vendor’s licensing agreement. 
• Compliance can also mean those efforts that ascertain if
organizations are abiding by both regulations of industry
and government legislation.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

❖Importance of Compliance

• In business, compliance is an important concern as there is an

increase in the number of regulations requiring companies to
be vigilant when it comes to maintaining a proper
understanding of the regulatory compliance requirements.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

 Compliance Risk
Management
• Managing compliance risk and passing a compliance audit
are the major responsibilities of the chief compliance officer in
an organization.

• Factors such as the:

o industry of the organization,
o private or public company,
o nature of data created,
o collected and stored

• determine the nature of the compliance audit of each company.

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA
• The protection of an organization can be assured on a whole
with regular regulatory compliance training programs
provided to IT staff members and business users.

• The guideline of a compliance training program differs

according to the industry a company is in, and the data
generated and used.

• The literal meaning of compliance is to formalize the act of

obeying an order, rule or request. In a corporate
environment, it refers to the state of being in agreement
with a list of guidelines or specifications that are widely
considered as a benchmark

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA

7. Auditing and Protection
❖ Intellectual Property (IP) Challenge

• Intellectual Property (IP) rights is still the biggest nightmare

in Big Data Analytics

• Rules:

o understand what IP is and what you need to protect.

o Prioritize protection
o Label (confidential information should be labeled)

Eng. N.F Thusabantu | MTech Big Data Analytics, INDIA