Download as pdf or txt
Download as pdf or txt
You are on page 1of 18

Data Science for

Cybersecurity Analysis
Julián Darío Miranda-Calle

DPhi Data Science


Bootcamp
DPhi Data Science
Bootcamp About me
Internal Auditor ISO 27001:2013, Cybersecurity Specialist,
Computer Science Engineer, and Electronics Engineer.
Currently, graduate and undergraduate professor at the
Faculty of Computer Science Engineering.

Scrum Master, developer and researcher in


cryptography, steganography, and steganalysis using
Data Science, Machine Learning and Deep Learning
techniques. Leader and Coach of the teams that will
Julián Darío participate in the XXXIV ACIS/REDIS National
Miranda-Calle Programming Contest 2020.

linkedin.com/in/juliandariomiranda/ 0000-0002-7580-2361
https://www.researchgate.net/profile/Julian_Miranda2
• Cybersecurity challenges DPhi Data Science
01 Introduction •

IoT services
Cyberattack practices
Bootcamp

• Key verbs
• DIKW Pyramid
02 Fundamentals •

Cyberattacks
CIA Triad
• Key concepts

Data Science
Agenda 03 Applied Data
Science
applications for
cybersecurity

Exploring
04 Use case cybersecurity
attacks

05 Conclusions •

Conclusions
Future work
Julián Darío Miranda-Calle
Cybersecurity Analysis
DPhi Data Science Julián Darío Miranda-Calle

Introduction
Bootcamp Cybersecurity Analysis

Cybersecurity challenges

Smart Washers Smart Toilets Smart Bicycle

Smart
Smart Blender Smart Shower Smart Bulb Lamppost
DPhi Data Science Julián Darío Miranda-Calle

Introduction
Bootcamp Cybersecurity Analysis

Cybersecurity challenges

Sexting Pharming Phishing

Social Engineering Vishing Smishing


Julián Darío Miranda-Calle
Cybersecurity Analysis
Introduction
Cybersecurity challenges

Hiring Home

Migrations Cybersecurity Industry


breaches and
challenges
Political Institutions

Financial Public
DPhi Data Science
Bootcamp
DPhi Data Science Julián Darío Miranda-Calle

Introduction
Bootcamp Cybersecurity Analysis

Cybersecurity challenges

Cybersecurity/Cybercrime/Cyber…

Big Data, AI, ML, DL, RL, FL & BI

Quantum Computing

Lifi/Zigbee/MQTT/AMQP/CoAP

Cloud Computing, IoT, IIoT & IoE


DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

Key verbs in cybersecurity

Knowledge
+Context Manage
Information
+Analysis Assure
+Interpretation
+Discussion

Data Protect
DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

DIKW Pyramid

DIKW (Data, Information, Knowledge, Wisdom) Pyramid


DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

Information properties

What makes the information so desirable to be accessed


without permission, kidnapped and/or deleted?

Information is …

Critical Sensitive
Information = Competitiveness
Information = Advantage

Valuable Information = Power


DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

CIA Triad

Ensure that the


information is
available for
access

Ensure that the Ensuring


information is that the
not accessed by senders are Ensure the
unauthorized who they deprivation
personnel say they are in the control
of access to
Ensure that the information
Information has
not been hijacked
during its Ensure monitoring
transmission of the processes for
accessing and
modifying
information
DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

Anatomy of a cybersecurity attack

01 03 05
Reconnaissance Gaining Access Covering Tracks
Steps taken to 02 Gain access to the 04 Conceal the success to
gather evidence on targets through the Keeping Access avoid detection by
Scanning
the targets to be vulnerabilities security professionals.
attacked. Gather more detected. Ensure the way
technical in-depth back to the target
information on the machines for
targets. further procedures.
DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Fundamentals Cybersecurity Analysis

Key concepts in cybersecurity

Any attempt to expose, alter, destroy, steal or gain unauthorized


Cyber attack access of an asset.

Source port Endpoint (between [0, 65535]) to a logical


connection that describes a service.
Logical port
Destination port

Set of rules to allow the transmission of information of two or


Protocol more entities in a system.

IPv4 Address
Numerical label assigned to devices connected
to a network that uses the Internet Protocol. IP Address
IPv6 Address
DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Applied Data Science Cybersecurity Analysis

How are data science practices applied in cybersecurity?

Risk assessment Digital Forensics Spam filtering Phishing emails


To search for traces Using NLP techniques Using NLP and
Clustering of to detect language Machine Learning
of file modification
common risks to components in spam techniques to unveil
and information
choose a treatment messages. phishing attempts.
violation

Event correlation
Data science in Network
cybersecurity: applications monitoring
Relation of multiple
Most of the techniques are build on top of Identification of
events to identify
EDA, Regression, Classification, supervised potential
common patterns
and unsupervised learning tasks cyberattacks
DPhi Data Science
Bootcamp Use case Julián Darío Miranda-Calle
Cybersecurity Analysis

Let’s analyze a dataset of real observations of


cybersecurity attacks to reveal some existing patterns
Source code and datasets available here
DPhi Data Science
Bootcamp Conclusions Julián Darío Miranda-Calle
Cybersecurity Analysis

The procedure explained in the use case can be extrapolated for all types of
cybersecurity studies in which cybersecurity attack data records are kept,
which can provide great information on what patterns occur and how to
effectively ensure services and servers

Data science allows to analyze the large volume of data that is generated
daily by IoT devices, industry, servers, home, among other contexts, to
propose new information security policies and countermeasures for
cybersecurity attacks to come.
DPhi Data Science Julián Darío Miranda-Calle
Bootcamp
Future work Cybersecurity Analysis

Current research

Artificial Intelligence applied to Digital Forensic Analysis:


A preliminary review. J. Miranda, J. Cano. 2019

LSB Steganography Detection in Monochromatic Still


J. Miranda, D. Parada. 2019 Images using ANN.

A study of the state of the art of NLP techniques for


Digital Forensics Analysis. Y. Reddy, J. Miranda. 2020

DL Techniques Applied against Steganalysis of Digital


M. Nabawi, J. Miranda. 2020 Images: A preliminary review.

Detection of LSB steganography in digital still images


using Convolutional Neural Networks. A. García, J. Miranda. 2020
DPhi Data Science
Bootcamp

Thanks!

juliandariomiranda@gmail.com /juliandariomiranda
Julián Darío Miranda
0000-0002-7580-2361 www.researchgate.net/profile/Julian_Miranda2

You might also like